Comparison of haemodynamic parameters between the high and low spinal block in young healthy patients

Background. For some surgical procedures a higher sensory block is needed. However, it is complicated by a higher incidence of hypotension, more bradycardia and nausea and a higher use of vasoactive drugs. In elderly and obstetric population complications have been attributed to the decrease in cardiac output and systemic vascular resistance, especially in a high block (above Th6). The aim of our study was to find the incidence of hypotension and bradycardia after a spinal anaesthesia in young, healthy patients. As young patients compensate more, we aimed to find the difference in haemodynamic variables between the group with a high and the group with a low spinal block and the underlying mechanisms of hypotension. Methods. In a prospective, randomized study 44 American Society of Anaesthesiologists (ASA) 1 patients scheduled for knee arthroscopy under spinal anaesthesia were randomly distributed to a high (group H) and a low (group L) spinal block group. In a group H patients were placed into horizontal, whereas in a group L in 15-degree anti-Trendelenburg position immediately after the spinal block. Haemodynamic parameters were measured continuously noninvasively from 10 min before to 25 min after the spinal block using the CNAPTM device with the LiDCORapid monitor. Results. The differences in haemodynamic parameters between the groups were not statistically significant at all measured times despite a significant difference in the spinal block level (18.5 vs 13.3 dermatomes above S5, p<0.001) and a significant difference in haemodynamic variables inside each group compared to the baseline value. With cardiac index (CI) as a dependent variable, a significant correlation between CI and stroke volume index (SVI) was found (β=0.849, p<0.001) and also between CI and heart rate (HR) (β=0.573, p<0.001). In group H the incidence of hypotension was 35%, whereas in group L it was 10%. The same difference was seen in the use of phenylephrine between the groups, however the difference was not significant. Conclusion. In our study it was found that in young, healthy patients there are no significant differences in haemodynamic parameters and in incidence of hypotension between a high and low spinal block. Young, healthy patients compensate a decrease in systemic vascular resistance caused by the spinal anaesthesia with a compensatory increase in CI resulting from an increase in SVI and HR. However, a trend towards less hypotension, less bradycardia and less frequent phenylephrine use in a low spinal block was noted.


INTRODUCTION
Spinal anaesthesia is a safe and routinely used anaesthetic technique, however it is related to various haemodynamic changes in the patient, most commonly hypotension.(1,2) The incidence of hypotension with a higher spinal block (Th7 or higher) appears to be as high as 60% or even more in elderly and obstetric population.(3)(4)(5) Hypotension may precede cardiac events (6) and increase 1-year postoperative mortality (7) therefore, many studies have focused on the prevention of hypotension due to spinal anaesthesia.(8) The underlying mechanism is the preganglionic sympathetic block and preserved or even increased parasympathetic nerve activity, (9) leading to a decreased systemic vascular resistance (SVR) and venodilatation, which causes a peripheral venous pooling of blood.The consequence is a lower inflow of venous blood into the heart, a decrease in cardiac output (CO) and finally a decrease in arterial blood pressure.(10,11) Besides this, in higher spinal block the loss of sympathetically mediated cardiac stimulation decreases heart rate and stroke volume leading to further decreases in CO.Hypotension during a high spinal anaesthesia is thus a result of decreased CO and SVR.(12) It has been shown that even if a decrease in CO was prevented (with application of colloids and/or phenylephrine), hypotension could not be prevented.(13) This shows that CO is unable to compensate a simultaneous decrease in SVR and that the decrease in SVR could be the dominant mechanism for hypotension during a high spinal anaesthesia.(14) On the other hand, there is considerably less hypotension in patients with a lower spinal block and the decreases in systolic arterial pressure (SAP) are modest (up to 20%).It is the consequence of the increase in sympathetic nervous system activity with a compensatory vasoconstriction in areas above the spinal block (thorax, upper limbs) (15)(16)(17) and a partial increase in CO, caused by the increases in heart rate (HR) and stroke volume (SV).(18) Besides hypotension, the sympathetic block also causes brady-cardia, nausea, vomiting, dysrhythmias and rarely cardiac arrest.(1) In previous studies it has been suggested that the incidence of hypotension depends on the level of spinal block and the age of the patient.(1, 3,4,19) The correlation between hypotension and an intrathecal dose of local anaesthetics was found with low doses resulting in less hypotension, vasopressor requirement, and nausea, presumingly because of a lower cephalic spread of local anaesthetic and a lower reduction of systemic vascular resistance.(20,22) Over 60 years ago, the difference in haemodynamic parameters between a high and low spinal anaesthesia was already noticed (the limit was Th4 dermatome), but no statistical analysis was performed.(23) In a study by Asehnoune et al (19) a significant difference in the change in CO was seen between two groups of patients with different sensory block levels: 6th and 8th thoracic vertebra.However, the study was performed on elderly and did not show a significant difference in the incidence of hypotension.Despite the above mentioned studies, the correlation between the spinal block level and the change in haemodynamic parameters has not been sufficiently investigated.The majority of studies on haemodynamic changes in spinal anaesthesia were performed on elderly (above 60 years of age) and parturient, however their haemodynamic status may be influenced by other factors.Elderly patients have higher decreases in systemic vascular resistance during spinal anaesthesia compared to young patients, (24,25) some authors consider it as the main mechanism of hypotension in elderly, (26) besides this, in the aged heart there is less of a compensatory increase in heart rate and contractility (decreased beta-adrenergic responsiveness), therefore a compensatory increase in cardiac output is smaller than in young adults or there is even a decrease in CO.On the other hand, the basal haemodynamic status of a healthy pregnant woman is characterized by a decrease in SVR, an increase in total blood volume and CO (27) and because of lower vascular tone, more blood volume is trapped in extremities.(28) Besides this, pregnant women are more susceptible to the effects of the sympathetic block.(29) Another limitation of many published studies is that the primary outcome of many studies was the incidence of hypotension, which was recognized by an intermittent non-invasive measurement of arterial blood pressure without advanced haemodynamic monitoring (CO and SVR).In this way, many episodes of hypo-tension were missed.(30) Based on the above literature, we came to a conclusion that there is a lack of studies that would examine the haemodynamic changes during spinal anaesthesia in healthy, younger subjects, which could provide the data for a better insight into the mechanisms of hypotension and would support the efforts to prevent it.This is the first study in young healthy nonobstetric patients that compares a high to low spinal block with the same dose of hyperbaric local anaesthetic using the continuous non-invasive recording of haemodynamic parameters.We hypothesized that different levels of the spinal block would result in significant differences in haemodynamic variables.

MATERIALS AND METHODS
In our prospective, randomized, singleblinded study 44 American Society of Anaesthesiologists (ASA) 1, patients aged between 18 and 40 years were included.Patients receiving spinal anaesthesia for knee arthroscopy between January 2014 and October 2015 were included in this study.The exclusion criteria included: chronic diseases (including peripheral arterial disease (PAD), Raynaud's syndrome and vascular surgery of the upper extremities; disorders of heart rhythm), drugs that could influence the patient's haemodynamic status, contraindications to spinal anaesthesia, patient's refusal to participate in the study, a history of allergy to local anaesthetic and a conversion to general anaesthesia.All patients were informed about the study and gave a written consent.Premedication was performed with 7.5mg of oral midazolam and patients were allowed to drink clear liquids up to 2 h before surgery.On arrival to the operating theatre, an 18-gauge intravenous cannula was inserted in a peripheral vein on the arm, not used for measurements.Standard monitoring (pulse oximetry, ECG, oscillometric upper-arm non-invasive blood pressure) was applied and connected with a Dräger Infinity Delta monitor (Drägerwerk AG & Co. KGaA, Lübeck, Germany).According to the risk/benefit ratio, the method of haemodynamic monitoring for a healthy patient population should be non-invasive (31) and according to availability at our institution the LiDCORapid monitor with CNAP was used.Therefore, the LiDC-ORapid v2 (LiDCO Ltd., Cambridge, United Kingdom) with CNAPTM monitoring system (CNSystems Medizintechnik AG, Graz, Austria) was applied on an index and middle finger on an arm without an intravenous cannula.The CNAPTM device develops an arterial waveform noninvasively by applying exterior pressure to the finger vessel wall keeping the blood volume of the finger arteries constant.The pressure in the cuff, which is needed to keep the volume constant, corresponds to arterial pressure.It is calibrated intermittently with a NIBP cuff.
The baseline measurements of haemodynamic parameters were performed after 10 minutes of lying in a supine position (systolic arterial pressure -SAP, mean arterial pressure -MAP, cardiac index -CI, systemic vascular resistance index -SVRI, stroke volume index -SVI and pulse pressure -PP).SVRI was calculated by the formula: SVRI = 80 x (MAP -RAP) / CI (by LiDCORapid), where RAP is the right atrial pressure.A RAP value was arbitrarily set to 7 mm Hg; namely, with non-invasive methods it was not possible to measure its real value.After the measurement, a spinal anaesthesia was performed in a sitting position at the L2-L3 or L3-4 interspace using a 26G atraumatic spinal needle with introducer (Atraucan, B. Braun Medical, Melsungen, Germany).In each patient 12,5 mg of hyperbaric bupivacaine 0.5% (Marcaine Spinal Heavy; Astra Zeneca, Lund, Sweden) was injected intrathecally over 30s and the patient was immediately placed in a supine position.After anaesthetic administration, the patients were randomized into 2 groups: group H (high spinal block) and group L (low spinal block) by the method of sealed envelopes (each envelope contained a number, generated by a computer and a paper saying "high" or "low block").All spinal blocks were performed by the same experienced anaesthesiologist.In group H, the patient was put in a supine position with the operating table in horizontal position after the induction of subarachnoid block, whereas in group L the table was tilted approximately 15 degrees in anti-Trendelenburg position for 10 minutes and then into horizontal supine position.The upper sensory level of spinal block was evaluated for 10 min in 2-minute intervals and in 5-minute intervals for another 20 min until there was no change in 3 consecutive readings.It was measured as a loss of cold sensation (using a sponge immersed in icecold alcohol).The haemodynamic variables were measured constantly with a recalibration each 5 min.A finger-cuff was switched from the index to middle finger and the opposite way each 20 min, as instructed by the CNAP company.An arm with the CNAP device was wrapped in warm swabs to prevent vasoconstriction because of low operating theatre temperatures and instructions were given to patients to keep the arm still.Time measurement started with the withdrawal of the spinal needle.Before the induction of spinal anaesthesia all patients received 2g of cefazolin in 100ml of saline over 15 min, after which an infusion of lactated Ringer's solution (5mL/ kg/h) was started.Hypotension was defined as SAP less than 80% of baseline -the value accepted by most investigators.(32)(33)(34) In case of hypotension, a bolus of 100 mcg of intravenous phenylephrine was given as a rescue medication with no additional fluid.If there was concomitant bradycardia (heart rate < 50 beats/min), a bolus of 0.5 mg atropine was given.The anaesthesiologist performing the block and measuring the sensory block level was blinded for the haemodynamic measurements.The monitoring of the patient as well as fluid and drug requirements were managed by another anaesthesiologist, blinded for the spinal block level.
The primary objective of our study was to determine the difference in CI values and in the incidence of hypotension between the patients with a high and the patients with a low spinal block.The decision to use the CI as the primary variable was based on the literature showing that inadequate CO results in reduced organ perfusion and an impaired microcirculation.(35) The secondary outcome was to observe the incidence of complications (nausea).

STATISTICAL ANALYSIS
The haemodynamic measurements were stored in the LiDCORapid monitor and downloaded as .csvfiles for each patient.The beat-to-beat data set was transformed into average values every 30 s.The data for the first 3 minutes after the spinal block were inconsistent because of a change in the patient's position from sitting to a supine, therefore those data were omitted.For the purpose of statistical analyses, dermatome levels were scored in sequence, starting at the 5th sacral segment (S5) with number 1. SPSS statistical program v.17.0 was used for the analysis of all data.The data were expressed as means, 95% confidence intervals for means or as a number of patients and percentage.For a normal distribution of the data, Kolmogorov-Smirnov test was used.The Paired Sample T-test was used to detect haemodynamic parameters' changes compared to the baseline values of the parameters, Independent Sample T-test to test the difference between the study groups and Multiple linear regression analysis to assess the relationship between the patient characteristics and baseline parameters with a change in CI, SAP, MAP, SVRI and HR.Chi-square test was used to compare the proportions of hypotension, nausea and the need for vasoactive drugs between the two groups.P < 0.05 was considered as statistically significant.According to preliminary data and the data from the literature, (19) a sample size of 20 patients per group was determined as sufficient to detect a 25% difference in cardiac index between high and low spinal anaesthesia group (SD 0,9 L/min/m2) with a power of 80% and the probability level of 0.05.To allow for a possible dropout rate of 10%, we needed 44 patients in this study.

RESULTS
Forty-four patients referred to knee arthroscopy were randomly distributed in 2 groups receiving high or low spinal anaesthesia.One patient showed excessive movement of the fingers due to anxiety, whereas 3 additional patients were excluded from data analyses because of problems with haemodynamic monitor calibrations.Therefore, 40 patients were included in the analysis with 20 in each study group.The variables in the majority of observed events and repeated measurements were distributed normally, the assumption that the data were distributed normally was confirmed (table 1).Both groups had comparable demographic data (age, BMI, gender and height) and the baseline values of haemodynamic parameters (CI, SAP, MAP, SVRI, PP and HR) prior to spinal anaesthesia (table 2).The level of sensory block was significantly higher in group H compared to group L (18.5 (95%CI 17.9-19.1)vs 13.3 (95%CI 12.7-13.9)dermatomes above S5, p<0.001) or converted into spinal segments 10th vs. 5th thoracic vertebral segment.The average time to reach the maximal level of sensory block was significantly longer in group H than in group L (12.8 (95%CI 11.4-14.2) vs. 10.9 (95%CI 10.2-11.6)min, p=0.026).All patients with hypotension received 1 bolus of 100 mcg of phenylephrine, resulting in 7 doses of phenylephrine in group H and 2 in group L, again the difference between the groups was not significant (p=0.064).A significant correlation between the phenylephrine consumption and sensory block level was found (Pearson correlation coefficient (ρ) = 0.367, p=0.020).There was a trend of higher incidence of nausea in group H, however the difference was not significant (table 5).No patient vomited during the study.

DISCUSSION
In this prospective randomized study, we found no significant difference in haemodynamic parameters between high and low spinal anaesthesia in young, healthy subjects, with which our hypothesis was disproved.However, there was a trend of advanced haemodynamic instability in patients with a high spinal block.Also, the differences in the incidence of hypotension were not statistically significant between the groups of patients.
To our knowledge, this is the first study that studied the influence of spinal anaesthesia level on haemodynamic variables in healthy, young, non-pregnant participants using non-invasive, continuous measurement method.Both study groups were similar in terms of age, gender, height, body mass index (BMI), time to maximal sensory level of spinal block and baseline values of haemodynamic parameters.
The level of sensory block was significantly higher in group H compared to group L, which was accomplished with the use of the same dose of hyperbaric bupivacaine (12,5mg) in both groups.The difference in the level of sensory block was achieved by the tilt of the operating table with a slight anti-Trendelenburg position in group L. It was shown before that with the positioning of the patient, different levels of sensory block could be achieved.(36)(37)(38) We decided for the dose of hyperbaric bupivacaine that is high enough to produce high spinal block (12,5mg), even in young patients.Namely, in young patients, the sensory block level after subarachnoid injection of hyperbaric local anaesthetic solution is usually 3-4 spinal segments lower than in elderly.(19,39,40) Hyperbaric bupivacaine was chosen as it is easier o get the reliable spinal block level.(41) In our patients (in group H) a higher spinal block was applied as would be required by the surgical procedure, however it was performed on ASA 1 population with a normal compensatory reserve, patients who could well tolerate haemodynamic shifts caused by high spinal anaesthesia.
In most studies, different levels of spinal block were achieved through different dosages of local anaesthetic (19)(20)(21)(22)42,43) and/or different baricity of local anaesthetics.(44) The conclusion of majority of studies was that with a lower dose of the local anaesthetic patients were more haemodynamically stable and that it was most probably the consequence of a different level of sympathetic block.(45,46) The drawback of lowering the doses (<5mg) was the increasing incidence of spinal block failures with pain, slower onset and shorter duration of the block and also the increasing rate of conversions into general anaesthesia.(42,44,47,48) In contrast to the above mentioned studies, Langesaeter et al. (20) showed that there was a difference in the incidence of hypotension with the same sensory block level, but a different local anaesthetic dosage.According to the above described findings, the focus of our study was the level of spinal block as the underlying factor for the difference in haemodynamic parameters during a spinal anaesthesia excluding different doses and/ or baricities of local anaesthetic.
A mean time to maximum level of sensory block was 12.8 minutes for group H and 10.9 minutes for group L. The timings are comparable to previous studies with a mean time of 15 minutes and a range 11-20 minutes.(49,50) The primary outcome of our study was the difference in CI value change between the group of patients with a high and the group with a low spinal block.CI is a primary determinant of global oxygen transport from the heart to the body, (51) therefore it was chosen as a primary outcome.In group H a significant increase in CI was found at all times.In group L there was a significant increase in CI values only 5 minutes after the block.Because of a lower block of sympathetic nerves in group L less compensatory increase in CI was necessary leading to insignificantly higher values of CI (above baseline) in this group 10 minutes or more after spinal anaesthesia.CI values in all patients were above baseline at any time after the spinal block.Also, in the study by Dyer et al. (18) it was shown that there is an increase in CI during a spinal anaesthesia, which could be caused by the increases in HR and SV.These findings were confirmed by our study, where a significant correlation between an increase in CI and an increase in SV and a correlation between an increase in the CI values and an increase in HR was found.The most probable reason for a CI increase is compensatory response to a significant decrease in a SVRI value (at the same time SVI increased significantly), but it could also be attributed to the change of the patient's position from sitting to supine.What is more, the haemodynamic curves showed the most prominent decrease in SVRI with a concomitant increase in CI in the first 5 minutes after the spinal block in group H and group L, with a peak effect after approximately 4 minutes.Similar findings were published by Langesaeter et al., (20)  The initial increase in HR 5 minutes after the block was most probably a compensatory response to a decrease in SVRI and it was probably also the consequence of a diminished parasympathetic activity to the heart.(9,18) Cook et al. (9) showed an initial increase in HR of 6-8% after the spinal block, which was higher than in our study (5%).In group L, a significant decrease in HR occurred after 15 min, which matched the time to the maximal spinal block and was the consequence of a sympathetic block.
All patients in our study received only a maintenance infusion of crystalloids to avoid the influence of fluids on haemodynamic parameters and to decrease the probability of urine retention.Despite this, the incidence of hypotension was low and it was treated with phenylephrine.The decision for the phenylephrine was based on the mechanism of hypotension (the decrease in SVR) and the safety and efficacy of the drug.(58) With the use of phenylephrine boluses, it was easier to keep the values of haemodynamic parameters closer to baseline values, (59) as it had a faster onset and could be more accurately titrated than other vasoactive drugs.(60) Besides hypotension, other common sideeffects related to hypotension were nausea, vomiting, and dizziness with incidences of around 25%. (61) In our study, the incidence of nausea in group H was 25%, whereas in group L it was 5%, the differ-ence between the groups was not significant.
Haemodynamic parameters in our study were measured using CNAPTM (continuous non-invasive arterial pressure) in combination with a LiDCO Rapid monitor.It was shown before that cardiac output and stroke volume were more valuable for detecting haemodynamic changes than conventional non-invasive monitoring (NIBP).(62) The CNAPTM device has been used as arterial pressure monitoring in patients with a spinal anaesthesia before (63) and it has detected more hypotensive episodes after a spinal anaesthesia than intermittent blood pressure devices (NIBP).Besides, a rigorous anaesthesia clinical trial showed real time estimates of arterial pressure by CNAP were comparable with an invasive arterial pressure, (64) a gold standard of invasive measurements.In spontaneous breathing, low-risk patients invasive monitoring using arterial line would not be ethically justified.(65) According to our experience, a CNAP device has failures in recordings and needs recalibration more often than the 20-min intervals, advised by the manufacturer.We also followed the advice from the company to synchronize a NIBP cuff of the main patient monitor with the CNAP device.
The PulseCO algorithm that was used in analysis of arterial pressure curve by LiD-CORapid monitor, has been validated and used in healthy, pregnant patients receiving spinal anaesthesia.(66,67) The absolute value of CI measured with this method has been controversial, however its value for monitoring the trends in CI values has been validated and accepted, also in spinal anaesthesia.(30,63,68) We were aware of the limitations of the method used in the assessment of absolute values of CI; also in SVRI calculations the mean atrial pressure was not measured but arbitrarily set to 7 mmHg.
The strengths of our study: after a rigorous literature review, it appears that this is the first prospective randomized study on spinal anaesthesia in young, healthy patients with two different levels of spinal block.
The study will certainly add to knowledge and will form the basis for future research in risk factors for hypotension during spinal anaesthesia and its prevention.
Our study had several limitations.The method used to measure haemodynamic parameters (CNAP) was susceptible to patient movement and low operating theatre air temperature.Besides this, the patient position was changed from sitting to supine after spinal anaesthesia influencing haemodynamic measurements in the first 5 minutes.Additionally, patients in group L were tilted anti-Trendelenburg position for the first 10 minutes probably causing more prominent compensatory vasoconstriction in the areas above the spinal block.The study involved a relatively small number of patients (it was powered to the difference in CI and not the other variables), which may impact the generalizability of the results.Also, it was only single blinded, as the researcher could not be blinded to block level, which adds to selection bias.

CONCLUSIONS
In young, healthy patients no significant difference was found in haemodynamic parameters between a group with a high and a group with a low level of sensory block using the same dose of hyperbaric bupivacaine.However, a trend towards less hypotension, less bradycardia and less frequent phenylephrine use in low spinal block was seen.This study proves that the main mechanism of hypotension in spinal anaesthesia is the decrease in systemic vascular resistance, however, young, healthy patients compensate much higher levels of spinal anaesthesia than other population.A possible strategy for reducing spinal anaesthesia induced hypotension and other haemodynamic deterioration according to our study still remains to minimize the peak block level to as low as possible for the planned procedure, which can be achieved only by adjusting the patient position.However, determining the most appropriate preventive measure and finding the risk factors for hypotension in young, healthy patients remains to be further studied.

Figure 1 .Figure 2 .Figure 3 .
Figure 1.Time course of percentage of change in cardiac index (CI) with respect to baseline in patients with a high (group H) and patients with a low (group L) sensory block.CI, cardiac index.*p < 0.05 with respect to baseline †p < 0.001 with respect to baseline

Table 1 .
The Kolmogorov-Smirnov and Shapiro-Wilk test of a normality of the distribution of the data in patients with high (group H) and patients with low (group L) sensory block.Cardiac index, SAP = systolic arterial pressure, MAP = mean arterial pressure, SVRI = systemic vascular resistance index, SVI = stroke volume index, PP = pulse pressure, HR = heart rate

Table 2 .
The presentation of demographic, general data and baseline values of haemodynamic parameters in patients with high (group H) and patients with low (group L) sensory block.Data are presented as means (95% confidence interval for means) or a number of patients (%).
BMI = body mass index, CI = Cardiac index, SAP = systolic arterial pressure, MAP = mean arterial pressure, SVRI = systemic vascular resistance index, SVI = stroke volume index, PP = pulse pressure, HR = heart rate Table 3. Haemodynamic parameters' values at Baseline, 5, 10, 15, 20, and 25 minutes after the spinal anaesthesia in patients with a high (group H) and patients with a low (group L) sensory block.The data are presented as means (95% confidence interval for means).Baseline 5

Table 4 .
Regression analysis with cardiac index (CI) as dependent variable and other haemodynamic parameters as independent variables.

Table 5 .
The frequency distribution of hypotension, bradycardia, atropine and ephedrine consumption in patients with a high (group H) and patients with a low (group L) sensory block.The data are presented as number of patients (%).