Acute respiratory acidosis does not increase plasma potassium in normokalaemic anaesthetized patients. A controlled randomized trial

European Journal of Anaesthesiology 2001, 18, 394–400

Acute respiratory acidosis does not increase plasma potassium in normokalaemic anaesthetized patients. A controlled randomized trial G. Natalini
1, V. Seramondi
, P. Fassini
, P. Foccoliy, C. Toninelliy, S. Cavalierey and A. Candiani



Department of Anaesthesia and Intensive Care, University of Brescia and yRespiratory Endoscopy and Laser Therapy Centre, ‘Spedali Civili’ Hospital, Brescia, Italy

Summary Background and objective Few and conflicting data are available regarding the changes of plasma potassium concentration during acute respiratory acidosis in human beings. This study compares the acute changes in plasma potassium concentration in acutely hypercapnic patients and in non-hypercapnic patients during general anaesthesia. Methods Thirty-three patients undergoing interventional rigid bronchoscopy were studied. Ventilation of the lungs was randomly conducted using either spontaneous-assisted ventilation or intermittent negative-pressure ventilation. All patients received the same anaesthetic protocol. Arterial blood gases and osmolality, and plasma concentrations of glucose, sodium, potassium and chloride were measured.

Introduction Acute acidosis can result in changes in plasma potassium concentration. A critical examination of the literature suggests that acute mineral acid acidosis consistently increases plasma potassium concentration; on the contrary, acute acid-organic acidosis is not associated with significant changes in plasma potassium [1,2]. With few exceptions, the available Accepted December 2000 Correspondence: G. Natalini, Terapia Intensiva Polifunzionale, Casa di Cura ‘Poliambulanza’, Via Bissolati 57, 25124 Brescia, Italy. 1 Present address: Department of Anaesthesia and Intensive care, Poliambulanza, Brescia, Italy.

394

Results Intraoperatively, PaCO2 was higher during spontaneous-assisted ventilation than during intermittent negative-pressure ventilation (9  1.8 vs. 5.4  1.2 kPa, P < 0.001) and the pH was also lower during spontaneous-assisted ventilation than during intermittent negative-pressure ventilation (7.24  0.07 vs. 7.4  0.08, P < 0.001). Plasma potassium concentration remained similar in both groups (3.8  0.2 mmol L1 with spontaneous-assisted ventilation vs. 3.7  0.4 mmol L1 with intermittent negativepressure ventilation). Conclusion Acute respiratory acidosis does not affect plasma potassium concentration. Keywords:

acidosis, respiratory; A N A E S T H E anaesthesia inhalation; D I A G N O S T I C POTASSIUM; RESPIRATORY SYSTEM, TECHNIQUES, bronchoscopy; V E N T I L A T O R S M E C H A N I C A L , ventilators, negative-pressure. SIA

ACIDOSIS,

GENERAL,

studies clearly demonstrate that acute respiratory acidosis results in an increment in plasma potassium lower than during mineral acid acidosis [1,3]. On this basis, hypoventilation during general anaesthesia (e.g. during spontaneous breathing) has been considered hazardous in hyperkalaemic patients [4]. Interventional rigid bronchoscopy under general anaesthesia offers an interesting experimental model to study the acute changes of acid-base balance and kalaemia developing due to hypoventilation. Mechanical ventilation of the lungs during interventional rigid bronchoscopy is limited by the absence of the cuffed tracheal tube. In our centre, spontaneous-assisted ventilation (SAV) has been the main modality of ventilation for 15 years [5,6]: this approach has been # 2001 European Academy of Anaesthesiology

Kalaemia during acute respiratory acidosis

shown to be safe in thousands of patients, although it involves moderate acute respiratory acidosis [5–8]. Recently, we have shown that intermittent negativepressure ventilation (INPV) can prevent hypercapnia during interventional rigid bronchoscopy [9,10]. In our randomized and controlled study, we compared the acute changes of plasma potassium concentration in acutely hypercapnic patients (SAV technique) and in non-hypercapnic patients (INPV) during interventional rigid bronchoscopy under general anaesthesia.

Methods The study was approved by the Ethical Committee of the Spedali Civili of Brescia and was carried out complying with the Declaration of Helsinki. Informed consent was obtained from all recruited patients. Thirty-three consecutive patients scheduled for interventional rigid bronchoscopy were recruited to the study. The characteristics of the patients are shown in Table 1. Patients with coronary artery disease or heart failure, life-threatening respiratory failure, cerebrovascular disease, diabetes mellitus, renal failure, drug therapy with digitalis or drugs modulating the autonomic nervous system were excluded from the study. Patients classified as ASA physical status III complained of incapacitating dyspnoea due to severe tracheobronchial stenosis of benign origin (postintubation, Wegener disease, etc.) and ASA IV patients underwent palliative treatment for the resection of endoluminal malignant neoplastic lesions. Patients were randomized to receive INPV (INPV group, 16 patients) or to be managed by the SAV technique (SAV group, 17 patients). Thirty minutes before the beginning of surgery the patients were invited to rest in bed and after 15 min an

arterial blood sample from a radial artery was collected; at the same time a venous catheter was inserted in the controlateral cephalic vein and a venous blood sample was collected. During the positioning of the venous catheter and the blood sampling, the patient was recommended to avoid any muscular activity. The arterial blood sample was collected with an arterial blood sampling kit consisting of a plastic syringe containing 50 units of lyophilized lithium heparin with a 22-gauge needle (Instrumentation Laboratory, Martell Medical, Temecula, CA, USA). Twenty minutes after induction of general anaesthesia, arterial and venous blood samples were collected at the same time. Preoperative and intraoperative blood samples were immediately analysed by the laboratory for arterial blood–gas analysis (ABL 300 Radiometer, Copenhagen, Denmark) and the venous plasma concentrations of glucose, sodium, potassium, chloride and osmolality were also determined (Monarch TM 2000, Chemistry System, Instrumentation Laboratory, Lexington MA, USA). Both groups received the same anaesthetic management. After 3 min of oxygen breathing, general anaesthesia was induced by administration of remifentanil 0.5–1 mg kg1 and propofol 2 mg kg1. Then lidocaine 4% (3 mg kg1) was sprayed into the trachea under laryngoscopic control and the rigid bronchoscope was introduced in the airway. Anaesthesia was maintained by a continuous infusion of propofol (6–8 mg kg1 h1) and remifentanil at the starting rate of 0.15 mg kg1 min1. The delivery rate of propofol were changed in cases of hypotension (systolic arterial pressure below 90 mmHg) and the remifentanil infusion rate was titrated to abolish coughing and, in the SAV group, to maintain residual spontaneous respiratory activity. During the study, fluid

Table 1. Patients’ characteristics

Number of patients Age (years) Weight (kg) Palliative treatment ASA grade II ASA grade III ASA grade IV

395

SAV

INPV

14 55  10 73  12 7 2 5 7

13 59  15 68  7 6 3 4 6

# 2001 European Academy of Anaesthesiology, European Journal of Anaesthesiology, 18, 394–400

P

0.419 0.203 0.853 0.839

396

G. Natalini et al.

administration was limited to isotonic sodium chloride solution. Patients requiring catecholamines or colloids to treat hypotension were removed from the study. In both groups, a continuous flow of oxygen was delivered through a lateral port of the rigid bronchoscope to maintain a peripheral oxygen saturation (SPO2) higher than 93%. Intermittent negative-pressure ventilation of the lungs was performed by a poncho-wrap connected to a negative-pressure ventilator (Emerson 33 C2, Emerson, Cambridge, MA, USA). INPV was started immediately after the insertion of the rigid bronchoscope. It was initially set to a sub-atmospheric (‘negative’) pressure (Pneg) of 25 hPa, a respiratory rate of 15 bpm and an inspiratory/expiratory ratio of 1:1. Spontaneous-assisted ventilation was applied after induction of anaesthesia when patients were able to breath spontaneously. If SPO2 was below 90% for more than 30 s, manual-assisted ventilation of the lungs with an anaesthesia bag at FiO2 ¼ 1 was delivered through the rigid bronchoscope until SPO2 reached 100%. Thereafter, the patients returned to spontaneous breathing. Moreover, irrespective of the SPO2 value, manual-assisted ventilation of the lungs was performed for 2 min in case of apnoea lasting for more than 5 min. These techniques are described elsewhere [9–11]. The electrocardiogram (ECG), heart rate (HR) and SPO2 were continuously displayed on a Propaq monitor (Welch Allyn Protocol, Beaverton, OR, USA). Non-invasive arterial pressure was measured every 2 min using the same monitor.

Statistical analysis Data are showed as mean  standard deviation. Comparison between SAV and INPV groups was performed by unpaired Student’s t-tests for parametric data and the Mann–Whitney U-test for non-parametric data. Differences in frequency were assessed using a w2 comparison. The relationship between changes in Hþ and Kþ was assessed by calculating the coefficient of correlation and the linear regression.

Results Three patients in each group were removed from the study because catecholamines and colloids were used to treat hypotension. Table 1 shows the main characteristics of the patients. Thirteen patients (48%) underwent palliative treatment of malignant endoluminal airway tumours. Nine patients (33%) were treated for tracheobronchial stenosis, two patients for papillomatosis and one for a carcinoma in situ. Finally, foreign bodies were removed in two patients. No differences between the groups were observed. The total dosage of propofol administered from the induction of anaesthesia to the blood sampling at T1 was 4.8  0.7 mg kg1 in the SAV group and 4.6  0.6 mg kg1 in the INPV group (P ¼ 0.435). The remifentanil infusion rate was 0.15  0.07 and 0.16  0.07 mg kg1 min1, respectively, during SAV and INPV (P ¼ 0.714). The amount of 0.9% NaCl delivered until the T1 samples was 0.497  0.065 L with SAV and 0.514  0.139 L with INPV (P ¼ 0.584). The preoperative

Table 2. Laboratory findings Preoperative SAV pH PaCO2 (kPa) PaO2 (kPa) Kþ (mmol L1) Glucose (mmol L1) Naþ (mmol L1) Cl– (mmol L1) Osmolality (mOsm kg1)

7.44  0.04 5.1  0.5 10  1.3 3.9  0.3 5.3  0.8 141  3.3 106  5.17 287.6  6

Intraoperative INPV 7.45  0.03 4.9  0.5 10.2  1.7 4.0  0.3 5.5  0.6 140  2.4 105  5.4 286.5  6.5

P

SAV

INPV

P

0.472 0.31 0.701 0.395 0.485

7.24  0.07 9  1.8 16.3  10.6 3.8  0.2 5.9  0.8

7.4  0.08 5.4  1.2 16  8.3 3.7  0.4 5.8  1.3

< 0.001 < 0.001 0.943 0.414 0.89

0.38 0.627 0.651

141  3.7 108  6.8 288  4.2

140  3 109  5.1 285  6.9

0.45 0.671 0.181

# 2001 European Academy of Anaesthesiology, European Journal of Anaesthesiology, 18, 394–400

Kalaemia during acute respiratory acidosis

397

Table 3. Cardiovascular parameters

1

Heart rate (T0) (beats min ) Heart rate (T1) (beats min1) Systolic pressure (T0) (mmHg) Systolic pressure (T1) (mmHg) Diastolic pressure (T0) (mmHg) Diastolic pressure (T1) (mmHg) SPO2 (T0) (%) SPO2 (T1) (%)

SAV

INPV

P

79  16 79  8 154  23 106  21 97  13 70  12 97  2 96  4

78  11 71  16 161  28 107  23 90  19 67  13 96  3 97  3

0.853 0.109 0.483 0.907 0.271 0.539 0.315 0.472

T0, preoperative value; T1, intraoperative value.

Fig. 1 The differences between the pre- and intraoperative Hþ (dHþ) and Kþ (dKþ) plasma concentrations (see text) are displayed graphically (linear regression and correlation: x ¼ 0.35 þ 0.007y, r ¼ 0.26, P ¼ 0.188). ^, SAV group; &, INPV group.

acid-base status, oxygenation, plasma electrolyte and glucose concentrations, and osmolality were similar in both groups (Table 2). During surgery, pH was significantly lower and PaCO2 and Hþ higher during SAV than during INPV. Thirteen patients in the SAV group had a PaCO2 higher than 6 kPa (93%), the other patient was normocapnic; in the INPV group only three patients (23%) were hypercapnic (PaCO2 > 6 kPa), eight (62%) were normocapnic and two (15%) hypocapnic (PaCO2 < 4.7 kPa). None of the other measured variables differed between groups. Pre- and intraoperative cardiovascular variables are shown in Table 3. No significant differences between groups were discovered. Figure 1 shows the relationship between the changes in Hþ and Kþ in all studied patients: no sig-

nificant correlation was observed (r ¼ 0.261, P ¼ 0.188). Moreover, the change in plasma potassium levels before and after anaesthesia was similar in both groups (0.19  0.26 mmol L1 in the SAV group and 0.32  0.36 mmol L1 in the INPV group, P ¼ 0.293). No major complications or adverse effects were observed either intraoperatively or perioperatively in the analysed patients.

Discussion Our study demonstrated that acute respiratory acidosis in anaesthetized patients did not affect plasma potassium concentration. Most previous studies have demonstrated that acute respiratory acidosis results in an increment of the plasma potassium concentration.

# 2001 European Academy of Anaesthesiology, European Journal of Anaesthesiology, 18, 394–400

398

G. Natalini et al.

This increase appears to be somewhat smaller than the increase during metabolic acidosis and strikingly lower than during mineral acid acidosis [1,3]. However, some studies have failed to show any correlation between acute respiratory acidosis and kalaemia [1]. Despite these conflicting findings, there are no recent studies published on this topic. There are practical implications for the correct approach to patients with acute respiratory acidosis. On the basis of the previous studies, hypoventilation (i.e. anaesthesia with spontaneous breathing) has been advised in hyperkalaemic patients [4]. From a theoretical point of view, the plasma potassium concentration should not increase during acute respiratory acidosis: an immediate greater increase in the hydrogen ion concentration takes place inside the cells and later outside the cells [12] and potassium should move into the cells in exchange for hydrogen ion. To our knowledge the present work is the second study on this issue conducted on anaesthetized human beings. The first study showed an increase in plasma potassium concentration related to the development of respiratory acidosis [13], but it was biased by important limitations and the results of their observation may be considered with some reservation. The role of blood insulin and glucose on the potassium partitioning between extra- and intracellular fluids is well recognized [1,14]. In the study of Finsterer and his colleagues, the groups differed for preoperative blood glucose concentration; intraoperative fluid therapy included the infusion of 5% glucose in water; blood glucose increased by about 8.33 mmol L1 intraoperatively: acute respiratory acidosis could not have been the only variable to explain the observed changes in the concentration of plasma potassium. Moreover, normocapnic patients received a fentanyl dosage greater (about 50%) than hypercapnic patients and no data on the cardiovascular response during anaesthesia were shown. Opioids reduce catecholamine release during the stress response to surgery [15,16], and catecholamines influence the potassium intraextracellular distribution [1]; a different adrenergic response could, at least partially, explain the differences in plasma potassium concentration between the study groups. Finally, hypercapnic patients showed a decrease in base excess greater than normocapnic patients; the metabolic change of the

acid-base status, as above described, could have affected the plasma potassium concentration. Our study, like most studies on the acute change of kalaemia due to acute respiratory acidosis, was conducted during anaesthesia [1]. The induction of anaesthesia may slightly reduce plasma potassium concentration [17]. Consequently, the results of these studies have to be considered with caution in awake patients. At the Spedali Civili Hospital, INPV during interventional rigid bronchoscopy is moving from the experimental application to the clinical setting. Nonetheless, SAV has been used for 15 years in more than 6000 patients and it is still widely used in clinical practice. INPV is generally recommended in patients with risk of poor tolerance to hypercapnia and acute acidosis (coronary artery disease or heart failure, life-threatening respiratory failure and cerebrovascular disease) [18]. Consequently, these patients were not randomized in this study. Moreover, we excluded from the study patients with conditions affecting the homeostasis (uptake, clearance and intraextracellular partition) of potassium (diabetes mellitus, renal failure, digitalis therapy or use of drugs modulating the autonomic nervous system). For the same reason patients requiring catecholamines to treat severe hypotension were excluded from the results [1]. The study population consisted mainly of high-risk patients but they were not bedridden and all were selfefficient. Patients recruited to the study were homogeneously distributed in both groups. As shown in previous studies [5–11], this study confirms that SAV is a safe technique during interventional rigid bronchoscopy, even when used for patients who are seriously ill. Sample timing is crucial in the evaluation of the results. A previous study showed that potassium changes after acute respiratory acidosis are timerelated, but in the first 30 min 65% of the final changes had occurred [13]. Our decision to take blood samples during surgery, 20 min after the induction of the anaesthesia, allowed us to evaluate only the immediate effect of the acid-base derangement on the plasma potassium concentration. Further investigation on the effects of acute hypercapnia was beyond the scope of this study. In this study, the total amount of anaesthetic and fluid was similar in both groups: it is therefore

# 2001 European Academy of Anaesthesiology, European Journal of Anaesthesiology, 18, 394–400

Kalaemia during acute respiratory acidosis

likely that any possible change in the plasma potassium concentration due to anaesthesia was similar. Furthermore, the effect of hypercapnia on the kalaemia should have made a difference between the groups. The only difference between groups in the pre- and intraoperative evaluation was the intraoperative level of PaCO2 and consequently of Hþ concentration, Kþ concentrations were constant in both groups. The degree of hypercapnia in the SAV group was from mild to moderate, whereas in the INPV group patients were hyper-, normo- or hypercapnic. If the whole study population was considered (Fig. 1), no correlation would be observed between the changes of Hþ and Kþ concentration. Furthermore, patients with extreme changes in Hþ did not reveal any tendency towards an increase or decrease in Kþ concentration. Acid-base balance and cardiovascular response in the INPV and the SAV groups were similar to values reported in previous studies [9–11]. We did not observe an evident adrenergic activation of the cardiovascular system during respiratory acidosis. For this purpose, the moderate degree of acidosis needs to be taken into consideration. Moreover, during interventional rigid bronchoscopy the opioid requirement in paralysed patients is lower that in non-paralysed patients [10]. We assumed that the required dosage to abolish coughing elicited by the rigid bronchoscope is higher than the analgesic dose for the interventional rigid bronchoscopy. An opioid dosage higher than the pure analgesic request could have a strong counteracting effect on the adrenergic system and explain the cardiovascular stability [14,15]. In conclusion, our study shows that acute respiratory acidosis does not acutely affect plasma potassium concentration in anaesthetized human beings. Accordingly, acute hypoventilation in hyperkalemic patients should be no more dangerous than in normokalemic patients. Further investigations should examine the effect of hypocapnia on plasma potassium concentration and the role of hyperventilation in mechanically ventilated patients with hyperkalaemia.

References 1 Adrogue´ HJ, Madias NE. Changes in plasma potassium concentration during acute acid-base disturbances. Am J Med 1981; 71: 456–467.

399

2 Adrogue´ HJ, Madias NE. Management of life-threatening acid-base disorders. (First part). New Engl J Med 1998; 338: 26–34. 3 Hishida A, Suzuki H, Ohishi K, Honda N. Roles of hormones in plasma potassium alteration in acute respiratory acidosis in dogs. Miner Electrolyte Metab 1992; 18: 56–60. 4 Roizen MF. Anesthetic implication of concurrent disease. In: Miller RD, ed. Anesthesia, 3rd edn. New York: Churchill Livingstone, 1990: 793–893. 5 Cavaliere S, Foccoli P, Farina PL. Nd-YAG laser bronchoscopy: a five-year experience with 1,396 applications in 1,000 patients. Chest 1988; 94: 15–21. 6 Cavaliere S, Venuta F, Foccoli P, Toninelli C, La Face B. Endoscopic treatment of malignant airway obstructions in 2,008 patients. Chest 1996; 110: 1536–1542. 7 Dumon JF, Shapshay S, Bourcereau J, et al. Principles for safety in application of neodymium-YAG laser in bronchology. Chest 1984; 86: 163–168. 8 Perrin G, Colt HG, Martin C, Mak MA, Dumon JF, Gouin F. Safety of interventional rigid bronchoscopy using intravenous anaesthesia and spontaneous assisted ventilation. A prospective study. Chest 1992; 102: 1526–1530. 9 Vitacca M, Natalini G, Cavaliere S, et al. Breathing pattern and arterial blood gases during Nd-YAG laser photoresection of endobronchial lesions under general anaesthesia: use of negative pressure ventilation. A preliminary study. Chest 1997; 112: 1466–1473. 10 Natalini G, Cavaliere S, Vitacca M, Amicucci G, Ambrosino N, Candiani A. Negative pressure ventilation vs. spontaneous assisted ventilation during rigid bronchoscopy. Acta Anaesthesiol Scand 1998; 42: 1063–1069. 11 Natalini G, Fassini P, Seramondi V, et al. Remifentanil vs. fentanyl during interventional rigid bronchoscopy under general anaesthesia in spontaneously breathing patients. Eur J Anaesth 1999; 16: 605–609. 12 Brown EB, Goott B. Intracellular hydrogen ion changes and potassium movement. Am J Physiol 1963; 204: 765–770. 13 Finsterer U, Luehr HG, Wirth AE. Effects of acute hypercapnia and hypocapnia on plasma and red cell potassium, blood lactate and base excess in man during anaesthesia. Acta Anaesthesiol Scand 1978; 22: 353–366. 14 DeFronzo RA, Sherwin RS, Dillingham M, Hendler R, Tamborlane WV, Felig P. Inflence of basal insulin and glucagon secretion on potassium and sodium metabolism. Studies with somatostatin in normal dogs and in normal and diabetic human beings. J Clin Invest 1978; 61: 472–479. 15 Vandermeulen EP. Controlling the stress response. In: White PF, ed. Textbook of Intravenous Anesthesia. Baltimore: Williams & Wilkins, 1997: 520–523. 16 Reisine T, Pasternak G. Opioid analgesics and antagonists. In: Hardman JG, Limbird L, Molinoff PB, Ruddon

# 2001 European Academy of Anaesthesiology, European Journal of Anaesthesiology, 18, 394–400

400

G. Natalini et al.

RW, Gilman AC, eds. Goodman and Gilman’s Pharmacological Basis of Therapeutics, 9th edn. New York: McGraw-Hill, 1996: 521–555. 17 Robson WL, Bayliss CE, Feldman R, et al. Evaluation of the effect of pentobarbitone anaesthesia on the plasma

potassium concentration in the rabbit and the dog. Can Anaesth Soc J 1981; 28: 210–216. 18 Feihl F, Perret C. Permissive hypercapnia. How permissive should we be? Am J Respir Crit Care Med 1994; 150: 1722–1737.

# 2001 European Academy of Anaesthesiology, European Journal of Anaesthesiology, 18, 394–400