Vives et al. BMC Anesthesiology (2019) 19:17 https://doi.org/10.1186/s12871-019-0682-1 RESEARCH ARTICLE Open Access Effects of bupivacaine or levobupivacaine on cerebral oxygenation during spinal anesthesia in elderly patients undergoing orthopedic surgery for hip fracture: a randomized controlled trial Roser Vives1* , Diana Fernandez-Galinski2, Francisca Gordo2, Alberto Izquierdo2, Joan C. Oliva3, Carmen Colilles2 and Caridad Pontes1 Abstract Background: Bupivacaine and levobupivacaine have similar pharmacokinetic and pharmacodynamic characteristics, and are used regularly in spinal anesthesia. Whether potential differences in their hemodynamic and anesthetic profiles could determine a differential risk of complications in elderly subjects, is controversial. The main objective was to compare the effects of intrathecally administered levobupivacaine (LB) versus bupivacaine (B), on regional cerebral O2 saturation during spinal anesthesia, cognitive status and neurological complications in elderly patients undergoing surgery for hip fracture. Methods: This was a randomized, controlled, single blind study. 58 patients aged 70 or older undergoing surgery for hip fracture with spinal anesthesia were allocated with a 1:1 ratio to receive LB or B, combined with fentanyl 15 μg, by intrathecal route. The primary outcome was the proportion of intraoperative time with regional cerebral desaturation (≥20% reduction in regional cerebral oxygen saturation from baseline), monitored by near –infrared spectroscopy. Secondary endpoints included hemodynamic parameters, level of sensory and motor block, changes in Short Portable Mental Status Questionnaire (SPMSQ), and neurological complications. Results: The mean percentage of intraoperative time with desaturation in the B group was 6.1% (SD: 17.5) and 4.7% (SD: 11.9) in the left and right hemisphere respectively; in the LB group the mean was 4.8% (SD: 11.4) in the left hemisphere and 2.4% (SD: 8.3) in the right one. No statistically significant differences were found between treatment groups. The level of sensory block at the start of surgery was lower for LB than for B (Th10 vs Th8, p:0.047) and motor block at 15 min was lower for LB (2.5 vs 3, p:0.009). No differences in postoperative SPMSQ were observed. Neurological complications such as confusional state, agitation or disorientation were reported in 50% of patients in the B group and 21.4% of patients in the LB group, p = 0.05. (Continued on next page) * Correspondence: Roser.vives@uab.cat 1 Departament de Farmacologia, de Terapèutica i de Toxicologia, UAB, Clinical Pharmacology Unit, Parc Taulí Hospital Universitari, Institut d’Investigació i Innovació Parc Taulí I3PT, Universitat Autònoma de Barcelona, Parc Taulí 1, 08028 Sabadell (Barcelona), Spain Full list of author information is available at the end of the article © The Author(s). 2019 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Vives et al. BMC Anesthesiology (2019) 19:17 Page 2 of 11 (Continued from previous page) Conclusions: No statistically significant differences in regional cerebral oxygen saturation or hemodynamic parameters were observed between both treatment groups. Bupivacaine and levobupivacaine differed in sensory and motor block achieved. While no differences were observed in cognitive impairment measured by the SPMSQ between treatment groups neurological complications reported by the physician were more frequent with bupivacaine. Trial registration: European Union Clinical Trials Register (EudraCT 2013–000846 -20) (April 9th, 2013). ClinicalTrials.gov (NCT01960543) (September 23rd, 2013). Keywords: Spinal anesthesia, Regional cerebral oxygen saturation, Bupivacaine, Levobupivacaine Background anesthetic profile between two different anesthetic agents Decline in cognitive function in elderly patients during used in spinal anesthesia. hospitalization has been, in part, attributed to sudden Currently, both B and LB are used interchangeably for changes in the usual environment of the patient, but also, spinal anesthesia in our hospital. Our working hypothesis to the acute morbidity or exacerbation of pre-existing con- was that cerebral desaturation would be higher in patients ditions. Geriatric patients suffering a hip fracture who treated with B than in those receiving LB, and that this undergo surgery are an especially frail population, because could be related to worse neurological outcomes. In order of concomitant morbidity, polypharmacy and bleeding. A to test the hypothesis, we designed the present pragmatic multi-factorial intervention aimed to preserve cognitive randomized clinical trial to compare the effect of B and function in this population during the acute hospital stay LB on regional cerebral oxygen saturation, during spinal has shown good results [1]. Spinal anesthesia would be part anesthesia in elderly patients undergoing surgery for hip of this multi-factorial interventions [1], although the evi- fracture repair, both administered by spinal route and ac- dence on the differential effects of spinal and general cording to routine clinical practice. We also compared the anesthesia on cognition is controversial. Shorter length of effects of both anesthetics on postoperative cognitive sta- hospitalization, which could result in fewer complications, tus and incidence of nervous system related adverse has been described for regional anesthesia, but it does not events. reduce mortality [2]. Intraoperative hypotension is an undesirable and frequent Methods effect of spinal anesthesia, which may complicate a pre- Design existing coronary heart disease, worsen previous mental This was a randomized, controlled, single-center, single impairment or precipitate a cerebral stroke. Different strat- blind trial conducted at Hospital de Sabadell. The proto- egies to maintain the hemodynamic stability of the patient col was approved by the Ethics Committee of the Centre include the choice of anesthetic drugs with the most favor- on March 2013 (CEIC Corporació Sanitaria Parc Taulí, able hemodynamic profile, and use of reduced doses of Sabadell, reference number 2013/015). The study was local anesthetic [3]. Racemic bupivacaine (B) and levobupi- registered at the European Union Clinical Trials Registry vacaine (LB) are closely related local anesthetic agents; (EudraCT 2013–000846 -20) on April 9th, 2013 and at controversial reports suggest that they have different ClinicalTrials.gov (NCT01960543) on September 23rd, anesthetic profiles, with B inducing higher blockade, more 2013. All patients received comprehensive information hypotension, and consequently, potentially more cerebral by the investigator on the study objectives, procedures desaturation [4, 5]. and risks and signed the informed consent form before Clinical monitoring of CNS is rarely performed during inclusion. Informed consent in front of a witness was spinal anesthesia. Instead, indirect parameters such as arter- allowed for patients unable to write or read due to physical ial pressure, heart rate and arterial O2 saturation are regis- impairment but with preserved cognitive function. tered [6]. Regional cerebral oxygen saturation (rCSO2) through near –infrared spectroscopy, a non invasive moni- Patients toring that reflects a balance between supply and consump- Patients aged 70 years and older requiring surgery for tion of O2 to the brain, has been proposed as a tool that hip fracture repair that were suitable to receive spinal may provide relevant information on CNS ischemic stress, block,were included. Prior to inclusion, patients under- and may be potentially useful to guide the clinical manage- went cognitive examination by means of the Short Port- ment of the patient during spinal anesthesia [7]. Thus, able Mental Status Questionnaire (SPMSQ) [8, 9] and rCSO2 might also be useful to assess the CNS conse- those with low scores (≥ 8 failures if elementary school, quences of the differences in the hemodynamic and ≥ 7 failures if higher education), were excluded. Patients Vives et al. BMC Anesthesiology (2019) 19:17 Page 3 of 11 were excluded if they had any contraindication to B or end of the surgery. The modified Bromage scale is used to LB (hypersensitivity to amide local anesthetics, severe measure the intensity of motor block based on the ability aortic stenosis, heart failure or coagulopathies) or if they of the patients to move their lower extremities (0 = no par- refused the regional anesthetic technique. alysis, 1 = able to move knees, unable to raise extended legs, 2 = able to flex ankles unable to flex knees, 3 = unable Methodology to flex ankles, knees or hips). After surgery, SB was evalu- A random list was generated by the clinical pharmacology ated every 30 min until regression to L5 level. unit using Winpepi V2.67 [10]. Sequentially numbered Regional cerebral oximetry was measured by near–infra- sealed opaque envelopes were produced containing the red spectroscopy using INVOS 5100 C, 7.00.0014 version treatment identity for each patient. The list and a copy of (Somanetics Inc., Troy, MI, 48084 - USA). A first measure the codes were kept in a file not accessible to the study of rSCO2 was performed at entry in the operating room. team. Inclusion was confirmed by the anesthetist, who A second measurement taken 2 min later was considered opened the first available sequentially numbered sealed as the reference baseline value for each hemisphere, and code to randomize the patient. The study was blind for values were recorded throughout the study. Regional cere- patients and for the anesthesiologists who evaluated the bral desaturation was defined as a 20% or larger reduction postoperative SPMSQ questionnaire. in rCSO2 with respect to the ipsilateral baseline value. Patients were not pre-medicated. Antihypertensive drugs Routine monitoring of SBP, mean arterial pressure (MAP), were discontinued on the day of surgery and restarted at HR and O2 saturation (O2 Sat) monitoring, volume of 48 h after surgery. Patients were randomly allocated (ratio crystalloid infused, intraoperative use of vasoconstrictor 1:1) to receive treatment with racemic bupivacaine (Bupiva- drugs, postoperative hemoglobin and the need for blood caine Braun 0.5%®, B. Braun Medical, S. A, Spain) or levo- transfusions, were also recorded. bupivacaine (Chirocane®, Abbvie Spain S.L.U, Spain), both The primary outcome measure was the mean propor- combined with fentanyl 15 μg, by intrathecal route. Both B tion of intra-operative time with regional cerebral oxygen and LB were isobaric. Spinal anesthesia was administered desaturation, defined as an rSCO2 relative reduction of through a 25 gauge spinal needle using the midline or para- 20% from baseline value. The area under the curve (AUC) medial approach. The B and LB dose administered was 7 for the intraoperative time below the desaturation thresh- mg if height was < 160 cm and 9 mg if height was ≥160 cm, old was also calculated (AUC rCSO2 reduced by ≥20% or when the surgical procedure was a hip prosthesis. Low from baseline). Similar calculations were done using an doses were used with the objective to maintain the hemo absolute value of 50% as a cutoff point (AUC rCSO2 < dynamic stability and minimize intra-operative hypoten- 50%). Minimum values (trough rCSO2) and time to first sion. Supplemental oxygen administration was started after saturation drop under the different cut off points were spinal injection. As per the protocol, if the anesthesia was also recorded. not sufficient to initiate surgery, midazolam 1 mg and pro- Postoperative cognitive impairment was evaluated after pofol 30 mg could be administered, and if not effective, 5 to 7 days, by changes in the SPMSQ scores from base- general anesthesia was induced and the patient was with- line. Any adverse events related to nervous system disor- drawn from the study. In the operating room patients ders, reported by the responsible physician within 30 days received 300 ml of Lactated Ringer’s over 20 min. Electro- after surgery were also recorded. cardiography, cuff for non-invasive blood pressure meas- urement, and pulse oximetry (Philips Intellivue MP70 Statistical methods Anesthesia) were used for routine monitoring. Hypotension A sample size of 28 patients per group was calculated (systolic blood pressure (SBP) < 100 mmHg) was treated assuming an expected clinically relevant difference between with iv phenylephrine 50–75 μg, and ephedrine 5 mg in treatments of 0.15 or higher on the mean proportion of in- case the hypotension did not revert after two doses of traoperative time with desaturation, with a standard devi- phenylephrine. Bradycardia (heart rate (HR) < 50 beats/ ation (SD) of 0.2, an alpha of 0.05 and 80% power. min) was treated with atropine 0.5–1 mg. Qualitative variables were described by the number of valid values and the frequency and percentage of each Outcomes and measurements category, and compared between groups using Fisher Patient characteristics such as ASA physical status, pa- exact tests. Quantitative variables were described by tient’s pathologies and treatments, type of fracture and measures of central tendency and dispersion, and com- surgery, SPMSQ and preoperative hemoglobin were pared between groups using a Student T test or Mann- recorded. In the operating room, and after spinal anesthe- Whitney’s test, as appropriate. A post-hoc bivariate sia, the level of sensory block (SB) was evaluated by pin- analysis using the above mentioned statistical tests was prick, and motor block (MB) by the modified Bromage performed to explore the relationship of selected vari- scale 15 min after the spinal injection, before and at the ables with intraoperative desaturation, postoperative Vives et al. BMC Anesthesiology (2019) 19:17 Page 4 of 11 cognitive impairment and neurological complications. admission to surgery was evenly distributed in both Variables showing statistical significance in the bivariate treatment groups. analysis were included in logistic regression multivari- No significant differences were observed between treat- ate analyses to explore predictors of regional cerebral ment groups in the proportion of intraoperative time with desaturation, cognitive impairment or neurological desaturation, nor in any of the secondary variables based complications. on the measurement of rCSO2 (Table 2). Overall, 46% of patients (13 B; 13 LB) presented a relative reduction in Results rCSO2 > 20% from baseline in either one or both hemi- From September 2013 to December 2014, 58 patients spheres; considering absolute values of rCSO2 below 50, meeting all the eligibility criteria were randomized. In 64.3% (B) and 53.6% (LB) showed desaturation at some two patients, the spinal anesthesia was not feasible and time point. Hemodynamic parameters (SBP, MAP and they were converted to general anesthesia; they were ex- HR) were comparable between the two groups, as well as cluded from the analysis. At the start of surgery, three the need for vasoconstrictor drugs, although a trend to a patients in the LB group needed midazolam + propofol lower SBP for patients was observed in B group (Table 3). and one patient in each group received midazolam The anesthetic profile for each treatment group is shown alone. Figure 1 displays the disposition of patients, and in Table 3. Intraoperative red blood cell transfusion was Table 1 shows the baseline characteristics of patients. required in 3 patients in the B group and in 4 patients in Most patients (57.1%) had at least 3 comorbidities at the LB group. The mean (SD) volume of crystalloid in- hospital admission. Table 1 shows the most frequent fused were 605.4 (321.6) in the B group and 589.3 (240.5) concomitant diseases per treatment group. 78.3% of pa- mL in the LB group (p = 0.83). tients were taking 3 or more different medications on a Postoperative SPMSQ scores were overall better than regular basis, and 53.6% more than 4. Table 1 shows the those observed at baseline (Table 4), and no statistically treatments most frequently used by patients per treat- significant differences were detected between treatment ment group. The majority of patients were operated on groups. Worsening in the SPMSQ score after surgery within 2 days after admission, and the time from was observed in 23.1% of patients in the B group, and in Fig. 1 Disposition of patients Vives et al. BMC Anesthesiology (2019) 19:17 Page 5 of 11 Table 1 Baseline characteristics Description of patients Bupivacaine (n = 28) Levobupivacaine (n = 28) N % N % Gender Female 19 67.9 24 85.7 ASA physical status 1 0 0 1 3.6 2 2 7.1 9 32.1 3 26 92.9 18 64.3 Education Elementary school 26 92.9 27 96.4 High school/University degree 2 7.1 1 3.6 Concomitant diseases Hypertension 18 64.3 18 64.3 Diabetes Mellitus 13 46.4 6 21.4 Dyslipidemia 7 25.0 8 28.6 Arrhythmia 7 25.0 5 17.9 Renal Failure 7 25.0 3 10.7 Cerebrovascular disease 1 4.2 4 16.7 Depression 1 3.6 6 21.4 Concomitant treatments Diuretic 14 50.0 11 39.3 ACE inhibitors 10 35.7 7 25.0 AR Blockers 6 21.4 8 28.6 Calcium antagonists 7 25.0 4 14.3 Beta-blockers 5 17.9 6 21.4 CNS drugs 19 67.9 25 89.3 Benzodiazepines 11 39.3 11 39.3 Antidepressants 8 28.6 12 42.9 Fracture type Pertrochanteric 17 60.7 11 39.3 Subtrochanteric 1 3.6 5 17.9 Femoral neck 8 28.6 7 25 Periprosthetic 1 3.6 2 7.1 Supracondylar 1 3.6 1 3.6 Distal diaphyseal 0 0 2 7.1 Surgery type Intramedullary internal fixation 16 57.1 14 50.0 Extramedullary internal fixation 5 17.9 8 28.6 Total hip arthroplasty 3 10.7 1 3.6 Cemented arthroplasty 4 14.3 3 10.7 Cementless hemiarthroplasty 0 0 2 7.1 Local anesthetic doses 7 mg 8 28.6 7 25.0 9 mg 20 71.4 21 75.0 Right rCSO2 lower than Left rCSO2 14 50.0 11 39.3 Vives et al. BMC Anesthesiology (2019) 19:17 Page 6 of 11 Table 1 Baseline characteristics (Continued) Description of patients Bupivacaine (n = 28) Levobupivacaine (n = 28) N % N % Mean (SD) n Mean (SD) N Age (years) 83.1 (5.5) 28 83.7 (6) 28 Weight (Kg) 66 (10.3) 28 67.6 (11.7) 28 Height (cm) 161 (7) 28 159.1 (6.3) 28 Baseline hemoglobin (g/L) 112.6 (19) 28 116.2 (18.1) 28 Duration on surgical intervention (min) 60.6 (24.7) 28 78.9 (35.5) 28 Baseline SBP (mmHg) 150.8 (24.7) 28 148.7 (21.8) 28 Baseline MAP (mmHg) 87.1 (14.2) 27 87.8 (14.4) 27 Baseline HR (bpm) 78.7 (11.7) 28 81.5 (11.5) 28 Baseline oxygen saturation (%) 92.4 (6.6) 28 93.9 (2.7) 28 Baseline rCSO2 (%) Left 59.8 (7.8) 28 59.7 (10.4) 28 Baseline rCSO2 (%) Right 58.8 (9.1) 28 60.1 (11.2) 28 ASA American Society of anesthesiologists, bpm beats per minute, HR heart rate, MBP mean blood pressure, SBP systolic blood pressure, SD standard deviation, rCSO2 regional cerebral oxygen saturation 28.6% of patients in the LB group. No differences in desaturation, increased by 61 fold in patients with base- changes from baseline in SPMSQ scores were observed line rCSO2 in the lowest quartile, and by 38 fold in those between treatment groups, even after adjusting by base- in the second quartile, as referred to the highest rCSO2 line SPMSQ scores. quartile (Table 5 and Fig. 2). Similarly, subjects in the Three patients from the B group died, and none in the lower quartile for trough intraoperative MAP had a 29 LB group (p = 0.24). Reasons were aspiration pneumonia fold increased risk of desaturation, as compared to the followed by heart failure, worsening of a pre-existing highest quartile (Table 5 and Fig. 2). No variable was chronic renal failure, and sepsis following a respiratory found to be significantly predictive of postoperative cog- infection, respectively. None of the deaths were consid- nitive impairment, defined as a worsening in the SPMSQ ered related to the study treatment by the investigators. score. In the multivariate analyses, only low SPMSQ Neurological complications were reported more often scores at baseline predicted neurological complications in patients in the B group (14, 50%) compared to the LB (OR = 1.58 for each additional incorrect answer in the group (6, 21.4%), p = 0.0496. Confusional state was re- baseline SPMSQ score, 95%CI: 1.14–2.21). ported in 10 patients (7 B, 3 LB), agitation in 3 patients (2 B, 1 LB), cognitive impairment in 2 patients (1 B, 1 Discussion LB) and disorientation in 2 patients (1 B, 1 LB); three Our study found no significant differences between B patients experienced agitation and disorientation (1 B), and LB in any of the endpoints based on regional cere- bradypsychia-dysartria (1 B) and encephalopathy related bral oxygen saturation. Cognitive impairment assess- to hypercalcemia (1 B). ment by SPMSQ did not show mental decline for most Baseline mean hemoglobin values were similar in both subjects, but patients receiving B had significantly more treatment groups (112.6 g/L and 116.2 g/L for B and LB reported neurological complications. The population in- respectively), and the drop in hemoglobin post-surgery cluded in our study was a very aged population (mean was comparable in both groups (106.2 g/L for B and age of 83.5 years, with 8 (14%) patients aged 90 or older). 110.4 g/L for LB). Anemia was reported as an adverse Patients receiving LB had lower sensory and motor event in 80.4% of patients, and was similar for B and LB blockades, as expected, and those receiving B had a groups, with 9 and 6 patients in B and LB groups re- trend towards lower SBP. This is consistent with the re- spectively, requiring postoperative transfusion (p = sults reported by Erdil et al. comparing B and LB (7.5 0.546). Urinary infections were reported in 4 and 5 pa- mg) in urologic procedures [5] and can be explained by tients, renal failure in 8 and 5 patients, and respiratory a higher density of LB compared with bupivacaine [11]. infections and/or respiratory failure in 3 and 3 patients rCSO2 is determined by a balance between cerebral in B and LB groups, respectively. oxygen supply and demand, and cerebral oxygen supply The multivariate exploratory analysis showed that in- is related to hemoglobin concentration, arterial O2 satur- traoperative desaturation was associated with low rCSO2 ation and blood volume; all of them are factors that can in the right hemisphere at baseline, so that the risk of be adversely affected during spinal anesthesia. Some Vives et al. BMC Anesthesiology (2019) 19:17 Page 7 of 11 Table 2 Intraoperative regional cerebral oxygen saturation with bupivacaine and levobupivacaine All patients Patients with any desaturation B LB B LB Desaturation > 20% below baseline Left hemisphere (n = 28) (n = 28) (n = 9) (n = 12) AUC (min %) 14.93 (42.60) 12.43 (30.86) ns 69.67 (72.28) 34.80 (44.63) ns Time below 20% from baseline (min) 5.57 (14.47) 5.21 (13.14) ns 17.36 (21.84) 12.14 (18.21) ns % of time with desaturation 6.07 (17.47) 4.82 (11.40) ns 18.9 (27.52) 11.25 (15.49) ns Time to first deep (min) – – ns 23.67 (22.31) 32.66 (30.27) ns Right hemisphere (n = 28) (n = 28) ns (n = 9) (n = 10) ns AUC (min %) 14.32 (35.34) 4.32 (14.98) ns 57.29 (52.43) 15.3 (26.05) ns Time below 20% from baseline (min) 4.74 (11.87) 2.44 (8.30) ns 14.75 (17.58) 6.84 (13.15) ns % of time with desaturation 4.01 (9.33) 2.16 (7.08) ns 14.02 (13.24) 6.04 (11.15) ns Time to first deep (min) – – – 29.28 (25.65) 45.97 (22.56) ns Desaturation < 50% (absolute values) Left hemisphere (n = 28) (n = 28) (n = 15) (n = 13) AUC (min %) 53.07 (108.82) 158.29 (440.69) ns 114.31 (138.06) 340.92 (607.88) ns Time below 20% from baseline (min) 12.82 (24.37) 16.59 (36.85) ns 23.93 (29.32) 35.75 (48.11) ns % of time with desaturation 11.62 (19.38) 13.65 (30.69) ns 21.68 (22.24) 29.41 (40.20) ns Time to first deep (min) – – ns 20.71 (30.77) 30.57 (35.09) ns Right hemisphere (n = 28) (n = 28) ns (n = 17) (n = 13) ns AUC (min %) 77.54 (152.84) 134.14 (422.71) ns 127.71 (180.6) 288.9 (594.63) ns Time below 20% from baseline (min) 15.84 (30.14) 13.32 (31.68) ns 26.09 (35.34) 28.68 (42.20) ns % of time with desaturation 13.08 (23.03) 11.12 (26.55) ns 21.54 (26,49) 23.94 (35.40) ns Time to first deep (min) – – – 28.80 (34.42) 19.12 (27.98) ns Lowest values rCSO2 (%) (n = 28) (n = 28) Left hemisphere 49.46 (8.59) 48.32 (11.71) ns Right hemisphere 48.18 (10.60) 49.86 (11.87) ns B Bupivacaine, LB Levobupivacaine, AUC area under the curve, rCSO2 regional cerebral oxygen saturation, ns not significant p value. Values expressed as mean (Standard Deviation) for all patients of the treatment group and for patients who presented desaturation of the measured hemisphere studies comparing spinal versus general anesthesia have groups, although differences in the anesthetic profile shown varying incidences of cerebral desaturation re- and rate of complications between B and LB were ob- lated to the anesthetic technique [12, 13]. During spinal served, suggesting that both anesthetic agents are not anesthesia in urologic surgery, Nishikawa observed a equivalent in efficacy. 10% reduction in rCSO2 in patients with higher levels of During surgery, we provided tailored prehydration and sensory block (>Th4.7) [14]. Also, in a randomized study supplementary oxygen, transfused red blood cells and in elderly patients undergoing abdominal surgery under used vasoconstrictor drugs (mainly phenylephrine) to general anesthesia, the use of rCSO2 monitoring avoided manage hypotension in every patient presenting SPB mea- cerebral desaturation and was associated with a reduced sures lower than 100mmHg [15]. Despite intensive man- incidence of postoperative cognitive decline [6]. agement, 46% of patients had at least one drop of rCSO2 We have predefined two cutoff points for assessment of 20% below the baseline value, in at least one cerebral of desaturation: a relative change ≥20% from baseline, hemisphere. Reference values of rCSO2 range from 60 to and an absolute saturation below 50%. We compared 80% [16]. At baseline our patients had mean rCSO2 values the percentage of time that patients spent below these around 60%, so that drops below 50% were easily and thresholds, and also the lowest values achieved. These quickly reached, while reductions of rCSO2 ≥ 20% from variables have not previously been reported for very eld- baseline were less frequent. These findings are consistent erly patients with spinal anesthesia, and in our study, we with the baseline rCSO2 of 63 ± 8% described by Casati did not observe any difference between the treatment et al. for relatively healthy patients aged around 70 [5], Vives et al. BMC Anesthesiology (2019) 19:17 Page 8 of 11 Table 3 Anesthetic and intraoperative hemodynamic profile of bupivacaine and levobupivacaine Anesthetic profile N Bupivacaine N Levobupivacaine Sensory Block (median – range) 15 min after SI 25 Th9 (Th4-Th12) 20 Th11 (Th5-TL2) p: 0.14 Start of surgery 23 Th8 (Th4-T12) 22 Th10 (Th5 – L2) p: 0.047* End of surgery 24 Th11 (Th5-L5) 21 Th11 (Th7- L5) p: 0.61 Duration of SB (min) (mean ± SD) 24 178.25 ± 44.89 22 175.18 ± 42.53 p: 0.81 Motor Block (median – range) 15 min after SI 25 3 (2–3) 20 2.5 (1–3) p: 0.009* Start of surgery 24 3 (2–3) 22 3 (1–3) p: 0.057 End of surgery 25 3 (0–3) 21 1 (0–3) p: 0.21 Duration of MB (min) (mean ± SD) 25 157.04 ± 41.71 23 141.65 ± 56.02 p: 0.29 Time (min) (Mean ± SD) SI to Start of surgery 25 21.32 ± 6.80 23 19.78 ± 7.45 p: 0.46 SI to End of surgery 25 80.75 ± 27.88 23 97.67 ± 35.97 p: 0.055 Surgery 28 60 ± 24.6 28 78.6 ± 35.4 p: 0.03* Hemodynamic profile Patients with hypotension (n/%) 28 19 / 67.8% 28 20 / 71.4% p: 0.99 Minimum values (Mean ± SD) Lowest SBP (mmHg) 28 84.0 ± 26.43 28 96.03 ± 21.31 p: 0.06 Lowest MAP (mmHg) 28 53.03 ± 14.11 28 56.96 ± 14.20 p: 0.31 Lowest HR (beats/min) 28 65.57 ± 10.66 28 65.57 ± 12.68 p: 0.52 Use of Vasoconstrictor drugs Phenylephrine n° patients / mean dose (ug) ± SD 28 17 / 258.82 ± 181.12 28 16 / 239.06 ± 153.83 p: 0.73 Ephedrine n° patients / mean dose (mg) ± SD 28 6 / 11.83 ± 5.19 28 7 / 15.57 ± 13.62 p: 0.52 Atropine 0.5 mg (n/%) 28 0/0.0% 28 3/10.7% p:0.24 SB sensory block, MB motor block, SI spinal injection, SBP systolic blood pressure, MAP mean arterial pressure, HR heart rate. SD standard deviation. Motor Block (0 = no paralysis, 1 = able to move knees, unable to raise extended legs, 2 = able to flex ankles unable to flex knees, 3 = unable to flex ankles, knees or hips) and Carlson et al. described that, awake elderly, with mar- the main hypothesis, and may be indicative of either the ginal rCSO2 levels (50–60%), had the greatest changes in lack of suitability of rCSO2 as a surrogate of cerebral oxy- rCSO2 during sleep that may affect subject’s cognition gen saturation during spinal anesthesia, or of a choice of [17]. Moreover, we observed a wide inter-patient variabil- thresholds unable to discriminate clinically relevant desat- ity in the number and duration of drops below our pro- uration. Future studies should consider including a more posed thresholds of rCSO2, especially in subjects already homogeneous population in terms of rCSO2 and the def- showing some degree of baseline desaturation. Such vari- inition of more stringent thresholds for the main ability may have compromised the power of the study for variables. Table 4 SPMSQ scores at baseline and after 5–7 days Baseline Post-op B LB B LB (n = 28) (n = 28) (n = 26) (n = 28) SPMSQ ns ns 0–2 errors 9 (32.1%) 13 (46.4%) 14 (53.8%) 16 (57.1%) 3–4 errors 12 (42.9% 9 (32.1%) 8 (30.8%) 8 (28.6%) 5–7 errors 7 (25.0%) 6 (21.4%) 4 (15.4%) 3 (10.1%) 8 or more errors 0 0 0 1 (3.7%) SPMSQ change, mean (SD) 0.50 (1.96) 0.32 (1.63) ns B Bupivacaine, LB Levobupivacaine, SPMSQ Short Portable Mental Status Questionnaire, SD standard deviation, ns not significant p value Vives et al. BMC Anesthesiology (2019) 19:17 Page 9 of 11 Fig. 2 Mean values and dispersion for (A) baseline right hemisphere rCSO2 and (B) minimum intra-operative MAP in patients with and without intraoperative desaturation. Footer: rCSO2: regional cerebral saturation of oxygen. MAP: Mean arterial pressure While in some studies intraoperative cerebral desatur- was considered, and this variable was only based on the ation has been found to be significantly associated with clinical observation of the physicians. This is a limitation an increased risk of cognitive decline [6, 18], a recent of the study and may be the reason for this finding. review did not show evidence of a relationship between Our post-hoc multivariate analyses also found signifi- a low rCSO2 and neurological complication [19]. Be- cant associations between preoperative SPMSQ scores sides, intrasubject differences in rCSO2 between L and R and NC, and the risk of intraoperative desaturation was during anesthesia has been also been suggested as a risk substantial and increasingly higher for patients in the factor for neurological impairment [20]. In our study, lower quartiles of baseline rCSO2 in the right hemi- the desaturation variables studied did not prove to be sphere. Preoperative SPMSQ has been widely reported useful as predictors of cognitive decline; however, we in the literature as a risk factor of cognitive decline [1], found that neurological complications were more fre- but low baseline rCSO2 in the right hemisphere as a risk quent with B. Such observation could be due to chance, factor for desaturation could be further explored as a but could also be due to the trends towards lower values potential marker of subjects in whom more preemptive of SBP and MAP in B group than in LB group [21], since hemodynamic management could be indicated. Salazar a higher sympathetic block may exaggerate hypotension et al. found that, in patients undergoing total knee re- after spinal anesthesia. In our study, no standardized placement, right rCSO2 values lower than left rCSO2 method to assess and report neurological complications was a factor associated with the presence of memory Table 5 Multivariate model for intraoperative desaturation risk N Mean SD OR (CI 95%) p value rCSO2 –Righta (overall) 56 59.41 10.128 0.020 1st quartile 15 47.53 7.954 61.1 (3.3 to 1127.9) 0.006 2nd quartile 14 55.79 1.626 37.6 (2.2 to 635.2) 0.012 3rd quartile 12 64.00 2.000 2.1 (0.4 to 12.1) 0.429 4th quartile 15 71.00 2.803 (Ref) MAPminb (overall) 54 55.00 14.17 0.145 1st quartile 13 36.23 7.33 28.7 (1.4 to 604.8) 0.031 2nd quartile 15 51.80 2.98 9.9 (0.8 to 122.3) 0.074 3rd quartile 13 59.92 1.89 4.3(0.4 to 51.3) 0.244 4th quartile 13 72.54 8.07 (Ref) Exp 0.085 0.045 OR Odds Ratio, CI Confidence interval, MAP Mean Arterial Pressure, SD Standard Deviation a Baseline regional Cerebral Saturation of Oxygen in Right Hemisphere; bMinimum value for Mean Arterial Blood pressure, intraoperative Vives et al. BMC Anesthesiology (2019) 19:17 Page 10 of 11 decline. We did not explore the asymmetry between left pressure; MB: Motor block; NC: Neurological complication; O2 Sat: O2 and right hemisphere saturation as a factor associated saturation; rCSO2: Regional cerebral oxygen saturation; SB: Sensory block; SD: Standard deviation; SPB: Systolic blood pressure; SPMSQ: Short Portable with cerebral desaturation or cognitive impairment, but Mental Status Questionnaire our finding that patients with lower right rCSO2 values present a higher risk of desaturation may be related to Acknowledgements Novella Caló, Carolina García-Bartolo, Cristian Hernández-Delgado, Anna this asymmetry. This finding should be further explored Valle-Beltran, Susana Fernández-Galinski, Marc Fradera. in future studies. Regarding the assessments of cognitive function based Funding This study had no external funding. on the SPMSQ, we did not observe differences either, since few patients worsened their scores after surgery. Availability of data and materials SPMSQ is a test easy to administrate and suitable for The datasets generated and/or analyzed during the current study will be very aged patients immobile due to the fracture, who available from the corresponding author on reasonable request. often have forgotten their glasses and who often had Authors’ contributions poor literacy. However, the first assessment was after a RV participated in protocol writing, statistical analysis, interpretation of results hip fracture and hospital admission, when the patients and manuscript writing. DFG conceived the study, participated in protocol writing, recruited and assessed patients and participated in the interpretation were in a estrange environment, in pain and anxious and of results and manuscript writing. FG recruited and assessed patients and were being treated with analgesic and sedating drugs. participated in the interpretation of results and reviewed the manuscript. AI Thus, this assessment may not represent the true cogni- recruited and assessed patients and participated in the interpretation of results and reviewed the manuscript. JCO did the statistical analysis and tive status of the patients before hospital admission; in reviewed the manuscript. CC recruited and assessed patients and turn, the second assessment was done after fracture participated in the interpretation of results and reviewed the manuscript. CP stabilization, control of pain and beginning of rehabilita- participated in protocol writing, statistical analysis, interpretation of results and manuscript writing. All authors read and approved the final manuscript. tion. Resolution of the previous painful conditions may explain why most patients presented better scores in Ethics approval and consent to participate comparison to baseline. A population undergoing elect- The study was approved by the Research Ethics Committee of our centre (Comité Ético de Investigación Clínica de la Corporació Sanitària Parc Taulí) ive surgery would be a more appropriate setting to check on March 2017. All participants were informed and asked for written the relationship between rCSO2 and cognitive decline. informed consent or oral informed consent in front of a witness before any Finally, 3 (10.3%) subjects died in the B group and none study procedure. in LB group. The reported rate of death in this particular Consent for publication population is 5.3% [2], and the observation is consistent Not applicable with these figures, considering the small sample size of our study that does not have power enough as to detect Competing interests The authors declare that they have no competing interests. statistically significant differences in mortality. Conclusions Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in This prospective, randomized, clinical trial in very eld- published maps and institutional affiliations. erly patients could not show differences in cerebral oxy- Author details genation between B and LB during spinal anesthesia, but 1 Departament de Farmacologia, de Terapèutica i de Toxicologia, UAB, Clinical lower sensory and motor blockade was seen with LB Pharmacology Unit, Parc Taulí Hospital Universitari, Institut d’Investigació i than with B, and a trend to lower MAP and SPB was ob- Innovació Parc Taulí I3PT, Universitat Autònoma de Barcelona, Parc Taulí 1, 08028 Sabadell (Barcelona), Spain. 2Anesthesiology Department, Parc Taulí served for B. While no differences were observed in cog- Hospital Universitari, Institut d’Investigació i Innovació Parc Taulí I3PT, nitive impairment measured by the SPMSQ between Universitat Autònoma de Barcelona, Parc Taulí 1, 08028 Sabadell (Barcelona), treatment groups, postoperative neurological complica- Spain. 3Statistics Unit, Parc Taulí Hospital Universitari, Institut d’Investigació i Innovació Parc Taulí I3PT, Universitat Autònoma de Barcelona, Parc Taulí 1, tions as reported by the physician were more frequent in 08028 Sabadell (Barcelona), Spain. B. An increased risk for intraoperative cerebral oxygen desaturation was observed for patients with lower trough Received: 30 January 2018 Accepted: 7 January 2019 intraoperative MAP and lower baseline rCSO2 values in the right hemisphere, and neurological complications References were associated to lower cognitive status at the pre- 1. Björkelund KB, Hommel A, Thorngren KG, Gustafson L, Larsson S, Lundberg operative assessment. Finally, rCSO2 was not found as a D. Reducing delirium in elderly patients with hip fracture a multi-factorial intervention study. Acta Anaesthesiol Scand. 2010;54:678–88. predictive factor for cognitive impairment or late neuro- 2. Neuman MD, Roenbaum PR, Ludwig JM, Zubizarreta JA, Silber JH. logical complications. Anesthesia technique, mortality and length of stay after hip fracture surgery. JAMA. 2014;311:2508–17. Abbreviations 3. Ben David B, Frankel R, Arzumonov T, Marchevsky Y, Volpin G. Minidose ASA: American Society of Anesthesiologists; AUC: Area under the curve; bupivacaine-fentanyl spinal anesthesia for surgical repair of hip fracture in B: Bupivacaine; HR: Heart rate; LB: Levobupivacaine; MAP: Mean arterial the aged. Anesthesiology. 2000;92:6–10. Vives et al. BMC Anesthesiology (2019) 19:17 Page 11 of 11 4. Gautier P, De Kock M, Huberty L, Demir T, Izydorczic M, Vanderick B. Comparison of the effects of intrathecal ropivacaine, levobupivacaine, and bupivacaine for caesarean section. Br J Anaesth. 2003;91:684–9. 5. Erdil F, Bulut S, Demirbilek S, Gedik E, Gulhas N, Ersoy MO. The effects of intrathecal levobupivacaine and bupivacaine in the elderly. Anesthesia. 2009;64:942–6. 6. Casati A, Fanelli G, Pietropaoli P, Proietti R, Tufano R, Danelli G, et al. Continuous monitoring of cerebral oxygen saturation in elderly patients undergoing abdominal surgery minimizes brain exposure to potential hypoxia. Anesth Analg. 2005;101:740–7. 7. Murkin JM, Arango M. Near infrared spectroscopy as an index of brain and tissue oxygenation. Br J Anaesth 2009;103 Suppl I: i3-i13. 8. Pfeiffer E. A short portable mental status questionnaire for the assessment of organic brain deficit in elderly patients. J Am Geriatr Soc. 1975;23:433–41. 9. Martinez de la Iglesia J, Dueñas Herrero R, Onis Vilches MC, Aguado Taberné C, Albert Colomer C, Luque Luque R. Cross cultural adaptation and validation of Pfeiffer’s test (short portable mental status questionnaire {SPMSQ}) to screen cognitive impairment in general population aged 65 or older. Med Clin (Barc). 2001;117:129–34. 10. WinpepiV2.67. At: http://www.brixtonhealth.com/pepi4windows.html, last accessed on August 14th, 2015. 11. McLeod GA. Density of spinal anesthetic solutions of bupivacaine, levobupivacaine, and ropivacaine with and without dextrose. Br J Anaesth. 2004;92:547–51. 12. Hoppenstein D, Zohar E, Ramaty E, Shabat S, Fredman B. The effects of general vs spinal anesthesia on frontal cerebral oxygen saturation in geriatric patients undergoing emergency surgical fixation of the neck of femur. J Clin Anesth. 2005;17:431–8. 13. Papadopoulus G, Karanikolas M, Liarmakopoulou A, Papathanakos G, Korre M, Beris A. Cerebral oximetry and cognitive dysfunction surgery for hip fractures: a prospective observational study. Open Orthop J. 2012;6:400–5. 14. Nishikawa K, Hagiwara R, Nakamura K, Ishizeki J, Kubo K, Saito S, et al. The effects of the extent of spinal block on the BIS score and regional cerebral oxygen saturation in elderly patients: a prospective, randomized and double-blinded study. J Clin Monit Comput. 2007;21:109–14. 15. Nakasuji M, HakSuh S, Nomura M, Nakamura M, Imanaka N, Tanaka M, et al. Hypotension from spinal anesthesia in patients aged greater than 80 years is due to a decrease in systemic vascular resistance. J Clin Anesth. 2012;24:201–6. 16. Tosh W, Patteril M. Cerebral Oxymetry. BJA Educ. 2016;16:417–21. 17. Carlson BW, Neelon VJ, Carlson JR, Hartman M, Bliwise DL. Cerebral oxygenation in wake and during sleep and its relationship to cognitive function in community- dwelling older adults without sleep disordered breathing. J Gerontol A Biol Sci Med Sci. 2011;66:150–6. 18. Slater JP, Guarino T, Stack J, Vinod K, Bustami RT, Brown JM 3rd, et al. Cerebral oxygen desaturation predicts cognitive decline and longer hospital stay after cardiac surgery. Ann Thorac Surg. 2009;87:36–45. 19. Zheng F, Sheinberg R, Yee MS, Ono M, Zheng Y, Hogue CW. Cerebral near infrared spectroscopy monitoring and neurologic outcomes in adult cardiac surgery patients: a systematic review. Anesth Analg. 2013;116:663–76. 20. Salazar F, Doñate M, Boget T, Bogdanovich A, Basora M, Torres F, et al. Relationship between intraoperative regional oxygen saturation trends and cognitive decline after total knee replacement: a post hoc analysis. BMC Anesthesiol. 2014;14:58. 21. The American Geriatrics Society Expert Panel on Postoperative Delirium in Older Adults. Postoperative Delirium in Older Adults: Best Practice Statement from the American Geriatrics Society. At: https://doi.org/10.1016/ j.jamcollsurg.2014.10.019, last accessed August 14th, 2015.