Local and systemic immune response to surgery under balanced anaesthesia in children.

We studied perioperative changes in the immune response and compared changes in the peripheral blood with those in the wound in 20 boys (0.5-3 years) during elective inguinal surgery under balanced anaesthesia. Blood samples were drawn before premedication, immediately, and 4 or 24 h after anaesthesia. Cells from the wound were harvested with the Cellstick device, removed from the wound 4 (n=10) or 24 h (n=10) after anaesthesia. We found decreased lymphocyte counts in the peripheral blood, increased percentages of activated T lymphocytes and B lymphocytes, and decreased percentages of total T lymphocytes, T helper cells and T cytotoxic cells. The percentages of T helper cells and B lymphocytes were lower in the wound than in blood. Mitogen-induced lymphocyte proliferative responses decreased. This study demonstrates perioperative changes in the immune response in children and, as a new finding, that immune effector cells in the blood and in the wound are in a dynamic balance.

Immune response in infants undergoing application of cast: comparison of halothane and balanced anesthesia.

PURPOSE: To study the immunological effects of two types of anesthesia on the immune response in infants during a minimally stressful surgical procedure. METHODS: The effects of inhalational halothane (halothane + N2O + O2, spontaneous breathing, n = 12) and conventional balanced anesthesia (thiopental + N2O + O2 + fentanyl + vecuronium, mechanical ventilation, n = 12) on immune function were measured in a crossover study in 12 infants undergoing application of casts to the lower extremity or hip joint. Leukocyte and differential counts, lymphocyte subpopulations, spontaneous lymphocyte proliferative responses as well as responses to phytohemagglutinin (PHA), concavalin A (ConA) and pokeweed mitogen (PWM), and serum cortisol concentration were measured before, immediately after and four hours after the end of anesthesia. RESULTS: Halothane anesthesia was associated with a higher percentage of T helper cells than conventional balanced anesthesia [47.1+/-1.8 (SEM)%, 48.1+/-2.3% and 50.7+/-1.9% before, immediately and four hours after anesthesia vs. 45.7+/-1.7%, 44.0+/-2.3% and 45.1+/-1.9%, respectively, by groups, P<0.05]. Leukocyte count and the percentages of activated T cells, natural killer cells and B cells showed similar alterations in both groups, and no alterations were observed in the percentages of T lymphocytes or T cytotoxic cells. Lymphocyte transformation response to PWM was decreased four hours after anesthesia in the halothane but not in the balanced anesthesia group. CONCLUSION: Anesthesia of short duration during minimal surgical stress alters lymphocyte subpopulations and lymphoproliferative responses in infants and, furthermore, halothane anesthesia and balanced anesthesia have different effects.

Contribution of non-neurologic disturbances in acute physiology to the prediction of intensive care outcome after head injury or non-traumatic intracranial haemorrhage.

OBJECTIVE: To study the additional contribution of non-neurologic disturbances in acute physiology and chronic health to the prediction of intensive care outcome in patients with head injury or non-traumatic intracranial haemorrhage. DESIGN: A nationwide study in Finland with prospectively collected data on all adult patients admitted to intensive care after head trauma or non-traumatic intracranial haemorrhage during a 14-month period. Two-thirds of the patients were randomly selected to derive predictive models, and the remaining one third constituted the validation sample. SETTING: A total of 25 medical and surgical ICUs in Finland (13 in tertiary referral centers). PATIENTS: 901 consecutive adult patients with head injury or non-traumatic intracranial haemorrhage. MEASUREMENTS AND RESULTS: Variables of the APACHE II including Glasgow Coma Score were collected at the time of ICU admission. Two predictive models were created to explain hospital mortality. The addition of variables describing acute physiology to a predictive model consisting of Glasgow Coma Score, age, diagnosis of head injury and the type of ICU admission did not increase its performance in discriminating between survivors and nonsurvivors, but the calibration accuracy of the predictive model especially at the high ranges of risk was improved. CONCLUSIONS: The non-neurologic disturbances in acute physiology have prognostic significance in the prediction of intensive care outcome in patients with head injury or non-traumatic intracerebral haemorrhage. The created predictive model may supplement clinical judgement of this patient group.

Prediction of outcome from intensive care after gastroenterologic emergency.

Prognostic factors determining the outcome from intensive care were studied in 952 patients admitted to 25 Finnish ICUs after gastroenterologic emergency. Logistic regression analysis was used to create predictive models based on the APACHE II-system. The models were constructed by using data from a random two-thirds of the study population and validated in the remaining independent one-third together with the original APACHE II-index. The Acute Physiology Score, age, and a pre-existing liver disease were the three most important determinants of outcome. The inclusion of the TISS score describing the intensity of treatment into a model did not enhance the accuracy of the prediction. Our models were better calibrated than the original APACHE II-equation when tested by the goodness-of-fit -statistics. These statistical models may help the clinicians to predict the outcome for an individual patient by providing them information about the relative impacts of predictive factors or about the probability of death. These probabilities should be interpreted cautiously, taking into account the limitations of statistical methods. This is especially important when assessing the highrisk patients. Their number in our study was too low for accurate outcome prediction.

Pharmacokinetics of glycopyrrolate in children.

STUDY OBJECTIVE: To investigate the pharmacokinetics of glycopyrrolate in children. DESIGN: Open study with three parallel groups. SETTING: Pediatric surgery department at a university hospital. PATIENTS: 26 healthy ASA physical status I children undergoing minor surgery. INTERVENTIONS: Patients were assigned to 1 of 3 groups: under 1 year of age (Group 1, n = 8), between 1 and 3 years of age (Group 2, n = 7), and over 3 years of age (Group 3, n = 11). Glycopyrrolate 5 micrograms/kg was given as a single intravenous (i.v.) injection before induction of general anesthesia. Blood samples (for determination of drug concentrations in plasma) were collected via venous cannula inserted into the contralateral antecubital vein. MEASUREMENTS AND MAIN RESULTS: ECG was observed continuously, blood pressure was measured with an automatic noninvasive device, and blood samples were taken just before and at 2, 4, 6, 10, 15, 30, 60, 120, 180, 240, 360, and 480 minutes after injection of glycopyrrolate. Glycopyrrolate concentrations in plasma were determined with a radioreceptor assay. The only significant difference in the pharmacokinetic parameters was the shortened elimination half-life in patients between 1 and 3 years of age. Glycopyrrolate 5 micrograms/kg i.v. did not cause any significant alterations in heart rate. CONCLUSIONS: There were no significant changes in the distribution volume or clearance of glycopyrrolate in children of different ages. The shortened elimination half-life in children between 1 and 3 years of age is of minor clinical importance.

Anticholinergic drugs: effects on oxygen consumption and energy expenditure.

Premedication has been shown to affect both oxygen consumption (VO2) and energy expenditure (EE). The metabolic responses to anticholinergic drugs have not been studied. In this study the effects of anticholinergic drugs on VO2 and EE (calculated from the measured rates of VO2 and carbon dioxide production [VCO2]: EE [kcal/d] = 3.581 x VO2 [L/d] + 1.448 x VCO2 [L/d] - 32.4) were measured in six healthy female volunteers. They were given intramuscular atropine (15 micrograms/kg), glycopyrrolate (8 micrograms/kg), scopolamine (8 micrograms/kg), and placebo in a random double-blind cross-over design. The consecutive sessions were at least 1 wk apart for each subject. VO2 and EE were measured using an indirect calorimetry (Deltatrac). Cardiovascular responses were assessed using standard noninvasive monitoring. Plasma drug concentrations were analyzed using a sensitive modification of radioreceptor assay. Subjective responses were measured with visual analog scale (VAS). Atropine and glycopyrrolate induced a significant increase in heart rate with a simultaneous decrease in pressure rate quotient (PRQ), while scopolamine caused a significant decrease in heart rate with a simultaneous increase in PRQ. Scopolamine significantly decreased both VO2 and EE, whereas glycopyrrolate increased VO2. Atropine had no significant effect on metabolic variables. Only scopolamine induced sedation in this study. In conclusion, atropine, glycopyrrolate, and scopolamine differ not only in their cardiovascular and central nervous system effects, but also in their effects on metabolism.

Systemic absorption and anticholinergic activity of topically applied tropicamide.

We studied the plasma levels and systemic anticholinergic activity of tropicamide after ocular administration in eight women. Two 40 microliters drops of 0.5% tropicamide were instilled into the lower cul-de-sac of one eye of the subjects and concentrations and respective muscarinic receptor occupancy of tropicamide in plasma were monitored using radioligand binding techniques. Tropicamide was rapidly absorbed systemically with the mean peak concentration in plasma being 2.8 +/- 1.7 ng/ml (mean +/- SD) at five minutes after instillation. Tropicamide disappeared rapidly from the systemic circulation: drug concentration in plasma was 0.46 +/- 0.51 ng/ml (mean +/- SD) at 60 minutes and below 240 pg/ml at 120 minutes after instillation. Tropicamide bound to muscarinic receptors of rat brain with an apparent equilibrium binding constant (Ki-value in plasma) 220 +/- 25 nM (mean +/- SD, n = 3). Tropicamide occupied maximally 8% of muscarinic receptors in plasma after ocular application. The low affinity of tropicamide for muscarinic receptors and its negligible receptor occupancy in plasma can explain the low incidence of systemic side-effects of tropicamide eyedrops.

Extent of beta 1- and beta 2-receptor occupancy in plasma assesses the antagonist activity of metoprolol, pindolol, and propranolol in the elderly.

We estimated antagonist activity of metoprolol, pindolol, and propranolol in elderly cardiovascular patients by determining the extent to which the drugs occupied rabbit lung beta 1- and rat reticulocyte beta 2-adrenoceptors in plasma samples during drug treatment. The randomized, double-blind, crossover study was carried out by administering twice daily 100 mg metoprolol, 5 mg pindolol, and 80 mg propranolol for 7 days to 20 hypertensive subjects with a mean age of about 70 years. A 2-week interval was kept between administration of the different regimens. Receptor occupancy was measured at 1 hour before and 2 hours after administration of the last dose of each regimen by adding rabbit lung beta 1- and rat reticulocyte beta 2-receptors to plasma samples and by labeling the receptors with a radiolabeled beta-antagonist, (-)-[3H]CGP-12177. The results and conclusions were the following: (a) The extent to which metoprolol, pindolol, and propranolol occupied rabbit lung beta 1- and rat reticulocyte beta 2-adrenoceptors in plasma samples estimated accurately the intensity of beta-receptor antagonism in the patients who did not tolerate physiological and pharmacological tests measuring the degree of beta 1- and beta 2-adrenoceptor blockade. (b) The mean beta 1- and beta 2-receptor occupancy of pindolol and propranolol varied between 76% and 99% during the treatments. The mean beta 1-receptor occupancy of the metoprolol regimen varied between 54% and 92%, and its beta 2-receptor occupancy varied between 6% and 38%. Thus the antagonist activity of the metoprolol regimen differed significantly from that of the other regimens (ANOVA for repeated measures, p < 0.05 and 0.001, for the beta 1- and beta 2-occupancy, respectively). (c) The extent of beta 1- and beta 2-receptor occupancy in plasma samples was in conformity with the literature on the intensity, selectivity, and duration of beta-blockade after similar drug doses. (d) The data on the receptor occupancy of beta-blocking drugs in plasma samples appear to be valuable in analyzing their effects, and it may be a method for optimizing drug therapy for aged cardiovascular patients.

Selectivity of acebutolol, atenolol, and metoprolol in healthy volunteers estimated by the extent the drugs occupy beta 2-receptors in the circulating plasma.

The selectivity of acebutolol, atenolol, and metoprolol in healthy volunteers was estimated by determining the extent to which the drugs occupied beta 1-receptors of rabbit lung and beta 2-receptors of rat reticulocytes in the circulating plasma after drug intake. This ex vivo method had the advantage of including all drug components contributing to the drug-receptor equilibrium in vivo and of excluding the factors regulating organ sensitivity to catecholamine stimulation. The oral doses of 400 mg acebutolol, 100 mg atenolol, and 100 mg metoprolol were administered to six healthy male volunteers using a double-blind, randomized, and cross-over study design. The three drugs occupied beta 1-receptors to a similar extent at 2 hours after drug intake. The receptor fraction occupied by metoprolol at 3 to 8 hours after drug intake was usually smaller, however (analysis of variance for repeated measures, P < .05) than that of the other drugs. Acebutolol occupied significantly larger fractions of beta 2-receptors (analysis of variance for repeated measures, P < .05) than did atenolol and metoprolol. Therefore, at an identical beta 1-receptor occupancy, the beta 2-receptor occupancy of acebutolol was larger than that of the other agents. Apparently, active metabolites decreased markedly the selectivity of acebutolol, but not that of metoprolol. The receptor occupancy of the agents was well in agreement with the literature concerning the selectivity, intensity, and time-course of drug actions after identical doses.(ABSTRACT TRUNCATED AT 250 WORDS)

Pharmacokinetics and related pharmacodynamics of anticholinergic drugs.

The pharmacokinetics and some pharmacodynamic properties of atropine, glycopyrrolate and scopolamine are reviewed. With the development of new analytical methods for drug determination, it is now possible to measure relatively low concentrations of these drugs in biological fluids and, consequently, some new kinetic data have been collected. Following intravenous administration, a fast disappearance from the circulation is observed and due to a high total clearance value their elimination phase half-lives vary from 1 to 4 h. All these agents are nonselective muscarinic receptor antagonists, but their actions on various organ systems with cholinergic innervation show considerable diversity. The cardiovascular effects are of short duration; other peripheral muscarinic effects and CNS effects can last up to 8 h or even longer. Differing from atropine and scopolamine, glycopyrrolate as a quaternary amine penetrates the biological membranes (blood-CNS, placental barriers) slowly and incompletely, making it the drug of choice for elderly patients with coexisting diseases and for obstetric use. Similarly, its oral absorption is slow and erratic, and hence it cannot be used as an oral premedicant. Atropine, scopolamine and glycopyrrolate have a definitely faster absorption rate, when injected into the deltoid muscle compared with administration into the gluteal or vastus lateralis muscles. There appear to be significant differences in the metabolism and renal excretion of these agents. Scopolamine is apparently excreted into the urine mainly as inactive metabolites, nearly half of the atropine dose administered is recovered in the urine as the parent drug or as active metabolites and about 80% of glycopyrrolate is excreted as unchanged drug or active metabolites.(ABSTRACT TRUNCATED AT 250 WORDS)

Absorption of ocular timolol: drug concentrations and beta-receptor binding activity in the aqueous humour of the treated and contralateral eye.

We studied the ocular and systemic absorption of 40 microliters of topical 0.5% timolol in 57 patients using radioligand binding techniques. The mean concentration of timolol in aqueous humour of the treated eye was 1.9 +/- 0.8 micrograms/ml 74 minutes after instillation of the drug. About 18 h after drug instillation the aqueous humour concentration of timolol was 105.5 +/- 60.9 ng/ml. Timolol was found in 15 (42%) contralateral eyes. Concentration of timolol in the contralateral eye increased from 0.04 +/- 0.08 ng/ml at 50 min to 0.3 +/- 0.2 ng/ml at 134 min and was 0.2 +/- 0.4 ng/ml at 18 h after instillation. Timolol concentrations in the aqueous humour of the treated eye appeared to be high enough to occupy beta 1- and beta 2-receptors completely (100%) at 74 min and at 18 h after drug instillation. Timolol concentrations in the contralateral eye were high enough to occupy up to 33.0 +/- 24.7% of the beta 2-receptors and up to 51.7 +/- 35.1% of beta 2-receptors. High drug concentrations and complete beta-receptor occupancy in the aqueous humour of the treated eye after topical timolol are in agreement with the long-lasting ocular hypotensive effects. The low drug concentrations and partial receptor occupancy in the contralateral eye may also be of some clinical significance.

Concentrations and antagonist activity of topically applied betaxolol in aqueous humour.

Aqueous humour concentrations and antagonist activity of betaxolol were studied after ocular administration in forty-five patients scheduled for cataract surgery. The patients were randomly divided into five groups and received 40 microliters of 0.5% betaxolol into the lower cul-de-sac of one eye. In groups I, II, III and IV the drug was instilled 5-6, 12, 24 and 48 h, respectively, before surgery, into the eye to be operated, and in group V 4 h before surgery into the contralateral eye. Aqueous humour samples were aspirated at the beginning of the operation. Aqueous humour concentrations of betaxolol were analyzed using a radioreceptor assay, and the ex-vivo-beta 1- and beta 2-receptor occupancies of betaxolol were calculated. The highest concentration of betaxolol in aqueous humour was found 5-6 hours after instillation of the drug. Topical betaxolol was found to stay in aqueous humour for 48 h, a much longer time than the recommended interval of dosage. Betaxolol beta 1-receptor occupancy was 99-95% during the study, but also beta 2-receptor occupancy was significant (52%) 24 h after instillation of the drug. Because receptor occupancy is the basis of antagonist activity, the role of beta 2-receptor blocking effect of betaxolol in lowering intraocular pressure cannot be excluded.

Beta 1- and beta 2-antagonist activity of topically applied betaxolol and timolol in the systemic circulation.

The beta 1- and beta 2-antagonist activity of betaxolol and timolol in the systemic circulation was studied ex-vivo after their ocular administration in thirty patients during cataract surgery. The patients received 40 microliters of 0.5% betaxolol or 0.25% timolol into the lower cul-de-sacs of both eyes. Blood samples were collected up to four h after instillation of the doses. Plasma concentrations of betaxolol and timolol were analyzed using a radioreceptor assay. The ex-vivo-beta 1-and beta 2-receptor occupancies corresponding drug plasma levels were calculated using radioligand binding techniques. The extent of beta 1-receptor occupancy of betaxolol in the systemic circulation was less than 20% and its beta 2-receptor occupancy was negligible. The extent of beta 1-receptor occupancy of timolol was about 65% and its beta 2-receptor occupancy about 80%. Because receptor occupancy is the basis of antagonist activity of beta-blocking agents, this study shows that the beta 1-antagonist activity of betaxolol in the systemic circulation is much less than that of timolol, and that its beta 2-antagonist activity is negligible. The study suggests that the reported side effects of betaxolol in patients with obstructive pulmonary diseases are not mediated via its beta 2-receptor blocking properties.

Pharmacokinetics of glycopyrronium in uraemic patients.

We studied the pharmacokinetics of glycopyrronium in 11 uraemic patients undergoing cadaveric renal transplantation and in seven ASA I control patients undergoing general surgery. Glycopyrronium 4 micrograms kg-1 was given i.v. before induction of anaesthesia. Blood and urine samples were collected for up to 24 h for measurement of glycopyrronium concentrations using a radioreceptor assay. Volume of distribution in the elimination phase (V beta) was similar in both groups, the elimination half-life (T1/2 beta) was longer (P < 0.05), area under the plasma concentration-time curve (AUC) larger (P < 0.01) and plasma clearance (CI) smaller (P < 0.01) in the uraemic patients. In 3 h, mean 0.7 (range 0-3)% and 50 (21-82)% of glycopyrronium was excreted in the urine in the uraemic and healthy patients, respectively (P < 0.001). The 24-h renal excretion was 7 (0-25)% in uraemic and 65 (30-99)% in control patients (P < 0.001). We conclude that the elimination of glycopyrronium is severely impaired in uraemic patients.

Kinetics of propiverine as assessed by radioreceptor assay in poor and extensive metabolizers of debrisoquine.

A pharmacokinetic study with 30 mg propiverine p.o. was performed in healthy volunteers (10 males, 6 females, age 36-56 years, body weight 55-100 kg, body height 162-184 cm, Broca index 0.96-1.19). 8 of them were poor and 8 extensive metabolizers of the debrisoquine type hydroxylation polymorphism. The total anticholinergic activity of the parent compound and active metabolites was measured with a radioreceptor assay calibrated with the metabolite M2. The affinity of this metabolite to the muscarinic receptors was similar to that of atropine. The urinary excretion of 3 major metabolites was determined with TLC and densitometry. Arterial blood pressure, heart rate, diameter of pupils, accommodation and parotic salivary flow were also measured. The concentrations of anticholinergic equivalents of propiverine were below 1 ng/ml of M2. 1.4-6.0% of the dose were excreted as N-oxidized metabolites into the urine. The poor and extensive metabolizers of debrisoquine did not differ significantly with regard to the concentration time behaviour of the active drug components, pattern of major metabolites, adverse drug reactions or any pharmacodynamic parameters measured.

The analysis of plasma kinetics and beta-receptor binding and -blocking activity of timolol following its small intravenous dose.

Plasma kinetics and beta-receptor blocking and -binding activity of timolol was studied in six healthy volunteers following its intravenous 0.25 mg dose. Timolol concentrations were measured using radioreceptor assay (RRA), blocking activity by comparing the dose ratios (DRs) of the infusion rates of isoprenaline required to increase heart rate by 25 bpm (I25) and binding activity by determining the extent to which timolol occupied beta 1-receptors of rabbit lung and beta 2-receptors of rat reticulocytes in undiluted plasma samples. Timolol was eliminated from plasma with a mean half-life for the elimination phase of 2.6 hours. The dose antagonized potently isoprenaline-induced tachycardia at least for four hours. The effect was excellently correlated with the estimated beta 2-receptor binding activity of timolol in the circulating plasma. In conclusion, the small intravenous timolol dose was eliminated from plasma by a fashion, which was very similar to its eighty-fold higher oral doses reported earlier in the literature. The 0.25 mg dose was of considerable systemic beta-receptor blocking and -binding activity, that may help to explain its reported side-effects following ocular drug administration. The extent to which beta-blocking agents occupy rabbit lung beta 1- and rat reticulocyte beta 2-receptors in the circulation appears to predict the intensity and selectivity of their beta-blocking effects in healthy volunteers.

Plasma and aqueous humour concentrations and systemic effects of topical betaxolol and timolol in man.

Plasma and aqueous humour concentrations and systemic effects of timolol and betaxolol were studied after ocular administration in 45 patients scheduled for extracapsular cataract extraction and intraocular lens implantation. The patients were divided randomly into three groups and received 40 microliters of either 0.5% betaxolol, 0.25% timolol or placebo into the lower cul-de-sacs of both eyes. Blood samples were collected over a period of 4 h and blood pressure and heart rate were monitored during the study. Aqueous humour samples were aspirated at the beginning of the operation. Plasma and aqueous humour concentrations of timolol and betaxolol were analyzed using a sensitive radioreceptor assay. The mean plasma concentrations of betaxolol were lower than those of timolol. The concentration of betaxolol in the aqueous humour was twice as high as the concentration of timolol. Both drugs produced a significant decrease in heart rate. In the timolol group a decrease in heart rate was found 15 min after drug administration, and in the betaxolol group after one hour.

The pharmacokinetics of mecillinam and pivmecillinam in pregnant and non-pregnant women.

1. The pharmacokinetics of parenteral mecillinam (n = 27) and oral pivmecillinam (n = 12) were studied in pregnant (n = 27) and non-pregnant (n = 12) subjects. 2. In early pregnancy (9-14 weeks of gestation) the mean peak plasma drug concentration (Cmax = 19 +/- 9 micrograms ml-1) after an intravenous injection of 200 mg mecillinam was significantly lower (P less than 0.05) and the volume of distribution (V = 49 +/- 20.1) significantly larger (P less than 0.05) than in non-pregnant subjects (Cmax = 35 +/- 18 micrograms ml-1, V = 29 +/- 12.1). In late pregnancy (39-40 weeks of gestation) the plasma mean peak concentration (Cmax = (29 +/- 14 micrograms ml-1) after parenteral administration of 200 mg mecillinam was slightly lower and the volume of distribution (V = 65 +/- 29.1, V = 0.9 +/- 0.4 l kg-1) significantly larger than that in non-pregnant subjects (V = 0.4 +/- 0.3 l kg-1). Also after oral administration of 200 mg pivmecillinam, equimolar to 136.5 mg mecillinam, the mean peak plasma concentration in pregnant subjects (Cmax = 1.8 +/- 1.2 micrograms ml-1) was slightly lower than that in non-pregnant subjects (Cmax = 1.7 +/- 1.2 micrograms ml-1). 3. The mean half-life of elimination after parenteral administration of mecillinam was significantly longer during both early (t1/2,Z = 133 +/- 38 min, P less than 0.05) and late pregnancy (t1/2,Z = 107 +/- 41 min, P less than 0.05) as compared with the non-pregnant state (t1/2,Z = 75 +/- 21 min).(ABSTRACT TRUNCATED AT 250 WORDS)

Tracheostomy in pediatric patients.

From 1978 to 1987, tracheostomy was performed on 33 children, 13 boys and 20 girls, with a male to female ratio of 0.65:1. The mean age at the time of tracheostomy was 726 days, 76% of the children being under the age of two years. The incidence of pediatric tracheostomy per hospital admissions was 0.05%. Subglottic stenosis (13 children) and respiratory distress syndrome with prolonged endotracheal ventilation (11 children) were the most common indications for tracheostomy. The mean duration of prolonged endotracheal intubation before tracheostomy was 64 days, and that of tracheostomy treatment 117 days. During the tracheostomy period, five children died, but only one death was related to tracheostomy. The total rate of complications was 30%. We emphasize the importance of strict indications for pediatric tracheostomy.