Print version ISSN 0034-7094
Rev. Bras. Anestesiol. vol.51 no.4 Campinas 2001
Predictors of early hypotension during spinal anesthesia*
Fatores de previsão de hipotensão arterial precoce em anestesia subaracnóidea
Factores de previsión de hipotensión arterial precoz en anestesia subaracnóidea
Getúlio Rodrigues de Oliveira Filho, TSA, M.D.I; Jorge Hamilton Soares Garcia, TSA, M.D.II; Ranulfo Goldschimidt, TSA, M.D.II; Adilson José Dal Mago, TSA, M.D.II; Marcos Aguiar Cordeiro, M.D.III; Felipe Ceccato, M.D.III
CET/SBA do SES/SC
IIInstrutor do CET/SBA do SES/SC
IIIME2 do CET/SBA do SES/SC
IVME1 do CET/SBA do SES/SC
BACKGROUND AND OBJECTIVES:
The identification of hypotension predicting factors during spinal anesthesia
may interfere in the choice of the technique or call for preventive measures.
This study aimed at identifying pre-anesthetic factors as independent hypotension
predictors during spinal anesthesia.
METHODS: Participated in this study 76 adult patients of both genders undergoing spinal anesthesia with 0.5% hyperbaric bupivacaine. The following parameters were collected: age, gender, weight, height, body mass index, systemic hypertension history, use of anti-hypertensive agents, preanesthetic medication, ankle and arm systolic and diastolic blood pressure, heart rate, pulse pressure, vascular overload, ankle-arm SBP index, bupivacaine doses, upper sensory block level and the lowest SBP measured at 2.5 minute intervals for 20 minutes. Patients were randomly allocated to receive oxygen (3 L.min-1). Hypotension was defined as systolic blood pressure below 80% of control or below 90 mmHg. Logistic regression was applied to identify independent hypotension predictors.
RESULTS: The following variables were independently associated to systolic hypotension: more than 45 years of age, female gender and upper sensory level above T7.
CONCLUSIONS: More than 45 years of age, female gender and upper sensory level above T7 were identified as independent predictors of early hypotension during spinal anesthesia.
Key words: ANESTHETIC TECHNIQUES, Regional: spinal block; COMPLICATIONS, arterial hypotension
JUSTIFICATIVA E OBJETIVOS:
A identificação de fatores de previsão de hipotensão arterial
durante bloqueios subaracnóideos pode interferir na escolha da técnica
ou na utilização de medidas preventivas. Este estudo avaliou fatores
pré-anestésicos como previsores independentes de hipotensão arterial
durante bloqueio subaracnóideo.
MÉTODO: Foram estudados 76 pacientes de ambos os sexos, submetidos à anestesia subaracnóidea com bupivacaína 0,5% hiperbárica. Foram coletados: idade, sexo, peso, altura, índice de massa corporal, história de hipertensão arterial, uso de drogas anti-hipertensivas, medicação pré-anestésica, pressão arterial sistólica (PAS) e diastólica (PAD) no braço e no tornozelo, freqüência cardíaca (FC), pressão de pulso arterial, índice de sobrecarga vascular, índice tornozelo/braço de PAS, doses de bupivacaína, nível superior do bloqueio sensitivo, e a menor PAS, medida a intervalos de 2,5 minutos até o vigésimo minuto. Oxigênio a 3 L.min-1 foi administrado, segundo sorteio. Hipotensão arterial foi definida como redução da PAS a valores inferiores a 80% do nível pré-anestésico ou PAS menor que 90 mmHg. Foi utilizada regressão logística para identificar as variáveis associadas com a ocorrência de hipotensão arterial.
RESULTADOS: Foram fatores de previsão de hipotensão arterial: idade maior que 45 anos, sexo feminino e nível superior do bloqueio sensitivo acima de T7.
CONCLUSÕES: Foram identificados fatores de previsão independentes de redução tensional sistólica acima de 20% dos valores pré-anestésicos: a idade acima de 45 anos, o sexo feminino e o nível superior do bloqueio sensitivo acima de T7.
Unitermos: COMPLICAÇÕES, hipotensão arterial; TÉCNICAS ANESTÉSICAS, Regional: subaracnóidea
JUSTIFICATIVA Y OBJETIVOS:
La identificación de factores de previsión de hipotensión arterial
durante bloqueos subaracnóideos puede interferir en la selección de
la técnica o en la utilización de medidas preventivas. Este estudio
evaluó factores pré-anestésicos como previsores independientes
de hipotensión arterial durante bloqueo subaracnóideo.
MÉTODO: Fueron estudiados 76 pacientes de ambos sexos, sometidos a anestesia subaracnóidea con bupivacaína 0,5% hiperbárica. Fueron colectados: edad, sexo, peso, altura, índice de masa corporal, historia de hipertensión arterial, uso de drogas anti-hipertensivas, medicación pré-anestésica, presión arterial sistólica (PAS) y diastólica (PAD) en el brazo y en el tobillo, frecuencia cardíaca (FC), presión de pulso arterial, índice de sobrecarga vascular, índice tobillo/brazo de PAS, dosis de bupivacaína, nivel superior del bloqueo sensitivo, y la menor PAS, medida a intervalos de 2,5 minutos hasta el vigésimo minuto. Oxigeno a 3 L.min-1 fue administrado, según sorteo. Hipotensión arterial fue definida como reducción de la PAS a valores inferiores a 80% del nivel pré-anestésico o PAS menor que 90 mmHg. Fue utilizada regresión logística para identificar las variables asociadas con la ocurrencia de hipotensión arterial.
RESULTADOS: Fueron factores de previsión de hipotensión arterial: edad mayor que 45 años, sexo femenino y nivel superior del bloqueo sensitivo arriba de T7.
CONCLUSIONES: Fueron identificados factores de previsión independientes de reducción tensional sistólica arriba de 20% de los valores pré-anestésicos la edad superior de 45 años, el sexo femenino y el nivel superior del bloqueo sensitivo arriba de T7.
The identification of patients with a high risk to develop hypotension during spinal anesthesia has already been object of a study resulting in a logistic model which considered sensory level above T6 and more than 50 years of age as independent predictors of hypotension during spinal anesthesia 1. Vascular overload index is a cardiovascular risk predictor in hypertensive patients and is correlated to the magnitude of systolic tensional decrease during general anesthesia and after propofol inducing doses 2,3. Ankle/arm systolic pressure index, in addition to predicting lower limb oclusive disease severity, is also decreased in hypertensive patients without peripheral arteriopathies, probably due to a change in blood flow velocity in the aorta and lower limb great vessels, caused by the stiffening of the arterial wall 4. By the same mechanism, vascular overload and ankle/arm systolic pressure indices are changed in hypertensive patients, although in opposite directions. Similar to vascular overload index, ankle/arm systolic pressure index is related to the magnitude of systolic tensional decrease during spinal anesthesia 5. Hypertension is or not considered a risk factor for hypotension during epidural and spinal anesthesia 6,7.
This study aimed at evaluating several preanesthetic factors as independent hypotension predictors during the first 20 minutes after spinal hyperbaric bupivacaine.
After the Hospital Governador Celso Ramos Medical Ethics Committee approval and their verbal consent, participated in this prospective study 76 patients of both genders, submitted to elective surgeries under spinal anesthesia with 0.5% hyperbaric bupivacaine. Patients were monitored with non invasive blood pressure, cardioscope (MC5) and pulse oximetry. The following preanesthetic data were collected: age, gender, weight, height, body mass index (BMI = weight.height-2), hypertension history, use of anti-hypertensive or cardioactive drugs, preanesthetic medication, systolic blood pressure (SBPcontrol), diastolic blood pressure (DBPcontrol), heart rate (HRcontrol), peripheral oxyhemoglobin saturation (SpO2control).
Lower limb systolic and diastolic blood pressures were also measured by an oscilometric method with the cuff right above the ankle. All blood pressure measurements were performed in the dominant limb (right, for all patients in this sample).
From preanesthetic hemodynamic data we calculated: arterial pulse pressure (PP = SBP - DBP), vascular overload index (VOI = 1.33 SBP - 0.33 DBP - 133), ankle/arm systolic pressure index (AMI = SBPlower limb/SBPupper limb).
All patients received 10 ml.kg-1 intravenous saline before spinal blockade.
Premedication, when administered, consisted of diazepam (0.15 mg.kg-1) or midazolam (0.03 to 0.05 mg.kg-1) associated or not to fentanyl (1 to 2 mg.kg-1) 5 to 10 minutes before anesthesia. Drugs were administered at the sole discretion of the anesthesiologist in charge.
SBP, DBP and HR measurements started soon after 0.5% hyperbaric bupivacaine spinal injection (M0) and measurements were repeated at 2.5-minute intervals for 20 minutes (M2.5 to M20).
The some patients received (3 L.min-1) oxygen during the observation period, according to a sealed envelope. Patients did not receive opioids or benzodiazepinics during the observation period.
Arterial hypotension was defined as SBP below 80% of control or below 90 mmHg. Ephedrine (5 to 10 mg) was administered to treat such episodes and oxygen under facial mask was used until SBP was back to more than 80% of preanesthetic values.
At the end of the observation period the following parameters were recorded: bupivacaine and ephedrine doses, upper sensory block level, confirmed by pinprick, and lowest systolic blood pressure (SBPmin) from which the maximum systolic tensional decrease percentage was calculated (DSBP = SBPmin - SBPcontrol/SBPcontrol).
Patients not meeting hypotension criteria during the observation period were allocated to Group N and those with hypotension to Group H.
Factors were submitted to individual analyses for comparisons between groups by Students t test (continuous variables) or chi-square test (categorical variables). Variables with p < 0.05 in individual analyses were considered independent variables in a logistic regression model. To make easier the logistic model, significant variables (gender, age and upper sensory block level) were split in two as from sample values median, as follows:
a. Gender: 1 - male, 2 - female;
b. Age: 1 - less than 45 years, 2 - equal to or more than 45 years;
c. Upper sensory block level: 1 - within or below T7, 2 - above T7.
Hemodynamic parameters, VOI, AAI and DSBP were compared between groups by Students t test for independent samples. Significance level was 5%.
Six patients were excluded from the study due to protocol noncompliance. From the remaining 70 patients, 30 (42.85%) had hypotension and were allocated to group H, while the remaining were allocated to group N. These groups differed in age, gender, weight, height, upper sensory block level and DSBP. There were no differences in BMI, SH, hypertension therapy, preanesthetic medication or oxygen therapy, bupivacaine doses, preanesthetic hemodynamic parameters, vascular overload index, pulse pressure or ankle/arm systolic blood pressure index (Table I, Table II and Table III).
Factors considered as independent hypotension predictors were: age equal to or above 45 years, female gender and upper sensory block level above T7 (Table IV). The model applied to the sample allowed for 77.5% of accurate classification of patients in Group N and 76.66% in Group H.
Hypotension during spinal block poses a risk for myocardial and brain ischemia 8. So, the identification of predicting factors for a higher hypotension risk during spinal anesthesia may help the anesthesiologist in his decision on the technique or on special monitoring measures and drug handling.
Logistic regression methods enable the building of predicting models of the probability of a certain event to occur in the presence of risk factors. Risk factors must be easily obtained and known before the blockade for the model to meet its preventive aim. In our study, routine data obtained during preanesthetic evaluations were tested as potential hypotension predictors. The observation period was limited to 20 minutes after blockade to prevent perioperative factors, such as bleeding and change in position, to affect the prevalence or intensity of systolic tensional decrease.
Treated hypertension is not related to a higher hypotension prevalence during spinal blocks 9 and this was confirmed by our study.
Vascular overload index, an alternative for arbitrary tensional levels in hypertension epidemiological studies 3, is related to the magnitude of systolic tensional decrease in patients submitted to general anesthesia 7. In a previous study 10, only hypertensive patients had a significant correlation between blood pressure decrease and vascular overload index after propofol-induced anesthesia. Ankle/arm systolic blood pressure index is correlated to previous hypertension 2 and to the magnitude of systolic tensional decrease after aorta clamping release in abdominal aorta surgeries 6. Such index, traditionally measured with doppler, has similar values when measured by the oscilometric method which is more widely available in the operating room, and that is why this was our method of choice 5. There have been no significant differences in vascular overload index or ankle/arm systolic blood pressure index between patients developing hypotension and normotensive patients, so they were not tested in the logistic model.
As in a different study 1, body mass index and pre-blockade administration of benzodiazepinics, associated or not to opioids, were not associated to hypotension. It was also not shown any preventive effect of oxygen administration on hypotension occurrence.
Independent hypotension predictors identified in our study were: age equal to or above 45 years, female gender and upper sensory block above T7.
Age has been related to hypotension. Several factors lead elderly people to be more prone to hypotension, such as a decrease in body water and poorer adaptation of the cardiovascular system to posture and volume changes 11.
In other study 12, female patients had 8.81 more chances of developing hypotension than males.
In our study, upper sensory block above T7 had 13 times more chances of developing hypotension. Sympathetic block extension, then, is still the factor more strongly related to spinal block hemodynamic changes 1.
The logistic model derived from our sample was robust in terms of accurately classifying 76.55% of Group H patients and 77.5% of Group N patients. So, the insertion of gender, age and upper sensory block level codes in the models formula:
P (hypotension) = 1/(1 + e - (-9.922 + 2.176 (gender) = 1.753(age) + (2.561(level))) , may predict the probability of a given patient to develop systolic hypotension during the first 20 minutes after anesthesia.
01. Tarkkila P, Isola J - A regression model for identifying patients at high risk of hypotension, bradicardia and nausea during spinal anesthesia. Acta Anaesthesiol Scand, 1992;36:554-558. [ Links ]
02. Hugue CJ, Safar ME, Aliefierakis MC - The ratio between ankle and brachial systolic pressure in patients with sustained uncomplicated essential hypertension. Clin Sci, 1988;74:179-182. [ Links ]
03. Franklin SS, Weber MA - Measuring hypertensive cardiovascular risk: the vascular overload concept. Am Heart J, 1994;128: 793-803. [ Links ]
04. Kelly R, Hayward C, Avolio A et al - Noninvasive determination of age-related changes in the human arterial pulse. Circulation, 1989;80:1652-1659. [ Links ]
05. Adiseshiah M, Cross FW, Belsham PA - Ankle blood pressure measures by automatic oscillotonometry: a comparison with Doppler pressure measurements. Ann R Coll Surg Engl, 1987;69:271-273. [ Links ]
06. Vretzakis G, Koutsias S, Grosomanidis V - A simple preoperative indicator of the haemodynamic changes during surgery of the abdominal aorta. Br J Anaesth, 1997;78(Suppl 1):A118. [ Links ]
07. Young C, Hilton A, Silbert K et al - Vascular overload index predicts intraoperative hypotension. Anesth Analg, 1996;82:S507. [ Links ]
08. Juelsgaard P, Sand NP, Felsby S et al - Perioperative myocardical ischaemia in patients undergoing surgery for fractured hip randomized to incremental spinal, single-dose spinal or general anaesthesia. Eur J Anaesthesiol, 1998;15:656-663. [ Links ]
09. Dagnino J, Prys-Roberts C - Studies of anaesthesia in relation to hypertension. VI: cardiovascular responses to extradural blockade of treated and untreated hypertensive patients. Br J Anaesth, 1984;56:1065-1073. [ Links ]
10. Oliveira Filho GR, Teixeira Filho N, Pederneiras SG et al - Índice de sobrecarga vascular e pressão arterial como indicadores de variação tensional sistólica após indução anestésica com propofol em pacientes normotensos e hipertensos. Rev Bras Anestesiol, 2000;50:269-272. [ Links ]
11. Critchley LA, Stuart JC, Short TG et al - Haemodynamic effects of subarachnoid block in elderly patients. Br J Anaesth, 1994;74:464-470. [ Links ]
12. Tarkkila PJ, Kaukinen S - Complications during spinal anesthesia: a prospective study. Reg Anesth, 1991;16:101-106. [ Links ]
Mail to: Submitted for publication Deciembre 4, 2000 * Received
from CET/SBA Integrado de Anestesiologia da SES/SC, Florianópolis, SC
Dr. Getúlio Rodrigues. de Oliveira Filho
Address: Rua Luiz Delfino 111/902
ZIP: 88015-360 City: Florianópolis, Brazil
Accepted for publication January 9, 2001
Submitted for publication Deciembre 4, 2000
* Received from CET/SBA Integrado de Anestesiologia da SES/SC, Florianópolis, SC