Over the last two decades, there has been a renewed interest in regional anesthesia for pediatric procedures. Several techniques have been extensively evaluated, many data are now available for virtually all pediatric age groups, and several text-books have been published [1-3] [4-8] [9] [10-20] [21] [22]
Methods
The study protocol, which included three separate forms, was sent to the 309 members of the ADARPEF. These three forms included an identification form, a data recording form, and a complication form.
The identification form was designed to describe the characteristics of the institution, i.e., administrative affiliation, number of beds, number of patients treated per year, geographical location, types of surgery performed, and the age ranges and pediatric experience of the anesthesiologists working in these institutions. This form also contained items requesting information concerning the pediatric cases, including the number of general anesthetics, number and types of regional blocks performed, and postanesthetic management of patients during the year 1992, i.e., the year preceding the prospective study.
The data recording form was to be completed every month, making a total of 12 forms. It included the following items: total number of pediatric cases with and without regional anesthesia, number and types of regional blocks, and type of sedation used for regional blocks, when applicable. The different block procedures were divided into three groups: 1) central blocks, including caudals, epidurals (sacral, Taylor, lumbar, and thoracic approaches), and spinals; 2) peripheral nerve blocks, including upper extremity and lower extremity nerve blocks, penile blocks, intercostal nerve blocks, and other blocks; 3) local anesthesias, including intravenous regional anesthesia (IVRA), field blocks, tracheal topical analgesia, intrapleural blocks, ear, nose, and throat anesthesias. For each of the three groups, the number of cases with placement of a catheter and administration of narcotics was requested. The patients were divided into seven age groups: 1) ex-premies less than 1 mo of age; 2) full-term infants less than 1 mo of age; 3) ex-premies aged 1-6 mo; 4) full-term infants aged 1-6 mo; 5) infants and toddlers aged 6 mo-3 yr; 6) children aged 6-12 yr; 7) adolescents more than 12 yr old.
The complication form had to be completed in the event of any adverse effects and one form per adverse effect was required. The items concerned the technique and equipment used, the dose given and time the drug was administered, a detailed description of the complication, where it occurred, management and patient outcome with documentation of neurological sequelae, and medicolegal implications if relevant.
The identification form was collected at the beginning of the study and was used to give each center an identification number. This number was given by a person not involved in the study and the data collection and statistical evaluation was also done by a person not involved in the study. The forms were collected every month and data were entered into the database. The complication forms were requested to be sent on a case-by-case basis after a complication took place and when the patient outcome and possible legal implications were known. These complication forms were kept confidential and neither the patient nor the institution could be identified. Statistical analysis of the results was performed by the University Laboratory of Statistics of the Faculty of Medicine of Marseilles using Microsoft Excel[R] (Microsoft Corp., Redmond, WA) and Epi-Info[R] (v. 5.0; Centers for Disease Control, Atlanta, GA) computer programs.
Results
Fifty-one percent of ADARPEF's members (164 of 309) participated in the prospective study. From May 1, 1993 to April 30, 1994, data from 85,412 pediatric anesthetics were prospectively collected. They included 61,003 pure general anesthetics and 24,409 anesthetics with a local or regional block procedure.
Results Obtained from the Identification Form
The 164 participants in the study worked in 38 institutions-33 in France, 4 in Belgium, and 1 in Italy. These institutions consisted of 28 university hospitals, 5 private hospitals, 4 nonuniversity hospitals, and 1 private medical clinic. The size of the institution varied from 100 to 1800 beds including 25 to 300 pediatric beds. There was a specific pediatrics department in 29 centers, whereas in 4, children and adults were treated in the same unit. There was no significant predominance of a special type of medical institution, either university, nonuniversity, or private practice hospitals, compared with the national distribution rate of institutions involved in the surgical management of pediatric patients. The centers that did not participate were evaluated and the reasons given by ADARPEF members were mainly: 1) they practiced in hospitals specialized in cardiac surgery, neurosurgery etc., where children usually accounted for less than 10% of the total patients; 2) they had no practice of regional anesthesia; 3) they had no spare time or data were not easily available in their institution. The potential bias due to the lack of information concerning the patients treated by these ADARPEF members not participating would probably not have significantly influenced the overall number of regional anesthetics whereas the overall number of general anesthetics would have been notably increased.
The anesthesiologists involved in the study were almost exclusively pediatric anesthesiologists (156 of 164); they performed 82,963 pediatric cases in 1992 (retrospective data), including 63,545 pure general anesthetics and 19,418 procedures with a regional block. According to the retrospective data of year 1992, the number of pediatric cases individually managed per year ranged from 100 to 1000. Data compilation of birth date and year of board certification displayed Gaussian distribution curves both in regard to age and experience, with a peak corresponding to anesthesiologists trained between 1980 and 1985, i.e., those in their mid-40s. Such a distribution is similar to that of the general population of anesthesiologists working in the same countries at the same time.
The different techniques of regional anesthesia were not uniformly used at all institutions. Peripheral nerve blocks were performed in only 28 institutions, whereas central blocks were performed in all 38 institutions. Placement of an epidural catheter was used in 30 centers and injection of epidural or spinal narcotics in 22. According to the retrospective data from 1992, postanesthetic care after central blocks varied considerably from one institution to another (Table 1 ). Epidural catheters were usually not kept in place more than 48 h (17 of 28 centers) and they were managed postoperatively in the intensive care unit (ICU) in 13 centers, in the recovery room postanesthesia care unit (PACU) only in 4 centers, in the recovery room/PACU then in the ward in 7 centers, or directly in the ward in 6 centers. Patients given epidural/spinal narcotics were monitored in the ICU (14 centers), PACU (4 centers), or ward (4 centers). Patients given peripheral nerve blocks were monitored in the recovery room/PACU in only 9 centers, in the recovery room/PACU then in the ward in 12 centers, and directly in the ward in 7 centers.
Table 1: Sites of Postanesthetic Management of Patients Given a Regional Block
Results from the 12 Data Record Forms
General Epidemiology of Regional Blocks in Children. The total number of pediatric anesthetics from May 1, 1993 to April 30, 1994 was 85,412 vs 82,963 in 1992, i.e., a 2.9% increase (Table 2 ). The overall number of pure general anesthetics was 61,003, compared with 63,545 in 1992 (-4%), and that of regional blocks was 24,409 compared with 19,418 in 1992 (+24.6%). The regional/general anesthesia ratio was 28.6% vs 23.4% in 1992. This ratio varied considerably from one institution to another, ranging from 4% to 100%. (In the latter case, data were obtained from a center that sent only three monthly forms and had a total of five pediatric anesthetics.)
Table 2: Epidemiology of Block Procedures in Children in 1992 and During the Prospective Study
Regional anesthetics were performed under light general anesthesia in 89% of patients, under sedation in 6%, and while fully conscious in 5%. The drugs used for anesthesia or for sedation were halogenated drugs (53%), benzodiazepines (21%), ketamine (11%), propofol (6%), and methohexital (4%). The drug administered is not reported in 5% of the procedures.
Central blocks (Table 3 ) accounted for 61.5% of all regional anesthetics and consisted mainly of caudals (50% of regional blocks). Epidurals (9.4% of regional procedures) were mainly performed at lumbar levels (72%) and only 6% were performed at thoracic levels. A catheter was inserted in 43% of epidural blocks, but in only 1% of caudal blocks and 0.6% of spinal blocks. Narcotics were injected in 5% of all central blocks, mainly epidurally i.e., in 19.5% of the epidurals, 2% of the caudals, and 9% of the spinals.
Table 3: Details of the Different Regional Block Procedures According to Patient's Age
Peripheral nerve blocks (Table 3 ) and local anesthetics accounted for 38% of all pediatric regional anesthesia. Local anesthesia included mainly spraying of the trachea with local anesthetics. There were only 69 IVRA.
Other blocks used were mainly penile blocks, whereas intercostal nerve and ilioinguinal/iliohypogastric nerve blocks were seldom performed. Extremity blocks (Table 3 ) represented only 6% of all regional procedures, two thirds of which were upper limb blocks (61% axillary blocks and 27% supraclavicular blocks). Lower extremity nerve blocks were mainly femoral nerve (46%) and sciatic nerve (38%) blocks. Blocks of nerves supplying the head and neck are not usually performed by anesthesiologists and thus very few were reported. These were included in the group entitled "other techniques" of the 524 trunk nerve blocks.
Age-Related Variations. The age-group distribution of regional block procedures is shown in Figure 1 . The general trend of the curve is that of a Gaussian distribution, with a peak corresponding to the group of patients aged 3-12 yr, which corresponds to the larger group of patients treated in pediatric institutions. The frequency of use of the different techniques was not uniform in the different age groups (Table 3 ). Caudals (12,111) accounted for 80% of central blocks in all age groups except in patients more than 12 yr of age, where it accounted for only 20%. In the latter group, intervertebral epidurals represented 70% of the central blocks, whereas in younger patients they accounted for significantly fewer than 20% of central blocks. Nearly 75% of all spinal anesthetics were performed in infants less than 6 mo old, including ex-premies (Table 3 ). Peripheral nerve blocks were used almost equally in all age groups. Conversely, the use of local anesthetic techniques was not uniform in all age groups; the peak frequency (2436 of 5306, i.e., 46%) was in patients aged 3-12 yr where local anesthesias represented 56% of all regional anesthetic techniques performed. IVRA was not commonly used (69 of 24,409, i.e., less than 0.3% of all regional anesthesias); none was performed in patients younger than 3 yr of age and 42 (60%) were performed in children more than 12 yr of age.
Figure 1: Frequency of use or regional block procedures according to patient's age.
Results Obtained from the Complication Forms
Twenty-five complication forms involving 24 patients were completed. They all involved central blocks: there were no reports involving either peripheral nerve blocks or local anesthesias. Two forms described postoperative problems with no clinical symptoms during continuous epidural analgesia. One child received antibiotics epidurally and the local anesthetic 0.125% bupivacaine was administered intravenously for 3 h; another case concerned a child weighing 12 kg who received 20 mL of 0.125% bupivacaine within 20 min due to malfunctioning of the electrical infusion pump. In neither case did the child have any clinical signs of toxicity, and in the first patient, it was the return of pain that led to recognition of syringe mix-up.
One patient had two complications after multiple attempts at performing a caudal block and two complication forms were completed. One form concerned a 17-yr-old male patient who should not have been included in the pediatric series; however, as anesthesia was performed by the pediatric team, this form was taken into account. The 23 reported complications (in 22 children) are shown in Table 4 . None resulted in death or neurological sequelae. In all cases the parents were informed of the complication, and although one family verbally complained of the anesthetic management, in no case was legal action taken against the practitioners or the institution.
Table 4: Details of the Reported Complications
Inadvertent dural puncture was the most frequent complication, especially during sacral epidural approaches, either via the caudal or the intervertebral sacral route. In four patients the epidural dose was injected intrathecally, thus resulting in total spinal anesthesia. In four patients, dural puncture was identified prior to injecting the local anesthetic; two patients complained of headache postoperatively, whereas the other two did not show any clinical symptom of inadvertent dural punctures. Intravascular injections were observed in six patients after a test dose with epinephrine had been performed; this test dose was considered uneventful in five patients but no precise criteria for "normality" was defined. This intravascular injection resulted in almost immediate seizures in two children, transient cardiac arrhythmias in two, and was subclinical in two. Two patients developed delayed cardiac arrhythmias 20 and 45 min, respectively, after injection, i.e., related to overdose rather than inadvertent intravenous injection: both were infants, both were ex-premies, and both received a mixture of 3.75 mg/kg of bupivacaine with 7.5 mg/kg of lidocaine. One patient experienced delayed postoperative apnea due to overdose of epidural morphine (100 micro g/kg in a 9-mo-old infant).
Three forms reported problems related to the technique used: there was one rectal puncture during a caudal approach, one knotting of the epidural catheter which was removed without breaking, and one case with delayed onset of the block. In two patients, short duration paresthesias, lasting 2 and 8 h, respectively, were reported postoperatively in areas supplied by the S-1 root, following the lumbar epidural approach at L3-4 and L4-5 levels, respectively. A causal link with the central block was not established, and nerve stretching due to intraoperative positioning could have been the cause, as early and complete recovery was seen in these patients. Finally, a patient was reported to have developed a round necrotic lesion at the puncture site of the caudal block when a solution containing epinephrine was used. In the latter too, the caudal cannot be entirely blamed as electric cautery was used and the patient was placed in the supine position on a heated water mattress. Three possible causes were postulated: 1) necrosis due to subcutaneous diffusion of the solution containing epinephrine; 2) heat burns due to the mattress; or 3) electrical burn favored by widespread diffusion of an excessive dose of skin cleansing solution which contained alcohol. Since the skin lesion disappeared within 3 days with no scar, the hypothesis of subcutaneous necrosis due to epinephrine seems improbable, whereas the possibility of a first degree burn seems more likely.
In 11 of 23 cases, the complication was directly related to the use of wrong equipment. The overall morbidity rate was 0.9 complication per 1000 regional anesthetics (23 of 24,409). In fact, no complications were reported after peripheral nerve blocks and local anesthesias (0 of 9396), all complications being reported after central blocks (Table 4 ); therefore, the morbidity rate for central blocks was 1.5 adverse effects per 1000 procedures (23 of 15,013). Considering the different techniques separately, the morbidity rate varied considerably from one technique to another (Table 4 ), the highest rates being observed with lumbar epidurals (4.6 per 1000) and intervertebral sacral epidurals (6.8 per 1000).
The male predominance of reported complications, 20 males versus 4 females, with information missing in one complication form, was still more marked (80%) even when taking into account the usual male predominance of patients operated in pediatric surgery (about 60%). With respect to the age groups (Table 5 ), complications appeared to be significantly more frequent in infants aged 1-6 mo, whereas the complication rate in the neonatal groups, either ex-premies or full-term neonates, was low, probably because these patients were managed by very experienced practitioners and with maximal precautions.
Table 5: Complication Rate in the Different Patient Age Groups
The complications occurred primarily in the operating room: 20 of 25 forms reported events occurring either in the operating room or in the PACU. The two patients who complained of postoperative paresthesias and the one who developed a sacral skin lesion were on the ward, whereas the other 20 (of 23) who presented with adverse effects with clinical symptoms were in the operating room at the time of diagnosis. In no case was the surgery postponed. In most cases (18 of 25), the anesthesiologist was experienced in the technique (used at least once a month); this technique was used for the particular operation systematically in 13 cases and commonly in 11 (information was missing in one form).
The complication forms included items aimed at providing information on diagnosis and management of the patient. Identification of symptoms and diagnosis were adequate in most of the cases (20 of 23). Acute management of the patients and emergency treatments when applicable were appropriate in all the cases. Conversely, mandatory laboratory examinations, especially the plasma dose of local anesthetics in case of seizures or cardiac arrhythmia, were not performed in any cases. Items concerning the attitude of the staff demonstrated that in all cases considerable help was offered by other staff members and surgeons. In no case did the adverse effect modify the indications for the technique in the relevant institution, but improvements of the protocol technique aiming at improving safety were made in all institutions, especially with regard to the equipment used. Finally, in no case did the complication result in legal proceedings; in one case only, the family expressed strong verbal dissatisfaction with care.
Discussion
The 12-month study protocol extending from May 1, 1993 to April 30, 1994 received enthusiastic acceptance from ADARPEF members, as more than 50% of them actively participated in the study [22]
Retrospective data from 1992 were used to display a general overview of the situation of pediatric anesthesia the year before the prospective study in the same institutions and can be used as a reference. The overall number of pediatric anesthetics increased by 2.9% during the study compared to data from 1992. Concomitantly, the number of pure general anesthetics decreased by 4%, whereas that of regional blocks increased by approximately 25%. Such an increase cannot be explained just by the inciting effect of the prospective protocol and is most likely related to the increased interest in regional block procedures. The majority of anesthesiologists used regional techniques as techniques of analgesia, not anesthesia, as most procedures (90%) were performed under light general anesthesia (with halogenated drugs in more than 50%) or sedation.
The postoperative management of the patients varied considerably from one institution to another (Table 1 ). Basically, the more sophisticated the technique the most intensive was the postanesthetic management of the patient, in PACU or ICU; however, there was no real consensus of attitude in this regard. The different regional techniques were not equally used in all institutions and in all age groups. Central blocks are the most frequently used regional anesthetics (61.5%), mainly represented by caudal anesthesia, which accounted for 81% of all central blocks and 50% of all pediatric regional anesthetics. Caudal anesthesia was the most commonly used technique in all age groups except that of children more than 12 years old (Figure 1 and Table 3 ). Intervertebral epidural anesthesia was less commonly used except in patients more than 12 years old; lumbar epidurals were the most frequently used (76%), followed by transsacral (12%), Taylor (10%), and, finally, thoracic epidural, which accounted for 6% of all epidurals. Unexpectedly, thoracic epidurals were mostly performed in infants less than six months old whereas, as anticipated, spinal anesthesias were first used in ex-premies and represented more than 2% of all regional anesthetics, i.e., twice as commonly used as either transsacral or Taylor epidural approaches. Epidural catheters and administration of narcotics along the neuraxis were not used in all institutions, and the overall number of catheter insertions was low (<5%). In spite of the lack of epidemiologic data, it seems that in recent years there has been a considerable decrease in catheter placement in children. The protocol study did not focus on the reasons why catheters were or were not used. We speculate two main reasons: 1) the rather high failure and morbidity rates of catheter especially compared with the ease and safety of single-shot procedures; 2) the increasing use of patient- or parent-controlled analgesia, which is well accepted (and usually preferred to continuous epidurals) by pediatric patients and provides high quality postoperative pain relief as the epidural of peripheral nerve block progressively recedes.
Compared with peripheral nerve blocks, central blocks appeared to be somewhat overused. We speculate that the reasons for this might be the former use of central blocks, especially caudals, and their reputation for ease and low morbidity rate. However, such an assumption is not confirmed in this study, and even if no severe complications were reported, all the adverse effects and complications were observed during and after central blocks including caudals (12 of 23 complications). Interestingly, virtually all potentially detrimental complications occurred in the operating room where adequate management could be (and had been) provided by a well trained anesthesiologist with adequate resuscitation equipment at hand: this was probably the main reason why no fatal outcome nor neurological sequelae were observed and should serve to remind that all regional blocks are anesthetic procedures, i.e., they have to be performed in an operating room environment by an experienced anesthesiologist [23,24]
One essential question raised by this prospective study concerns the relevance of the reported complications. Whereas the occurrence of fatalities could not have been hidden, there is a possibility that all the complications were not reported. However, due to the intensity of the ADARPEF campaign, and because, for each institution, several different participants were involved at the same time in the collection and control of data entries, it is improbable that a significant number of complicated cases were lost. Also, it would have been interesting to obtain data concerning the morbidity and morality rate of the pediatric patients operated during pure general anesthesia during the same period. However, the questionnaires did not include relevant items as the study protocol focused specifically on regional anesthesia cases and their complications.
One essential finding of this study was that in 11 cases improper equipment had been used, e.g., inappropriate or oversized needle, excessive length of catheter introduced into the epidural space. With proper equipment and technique these 11 complications could have been avoided. Also, two patients received an overdose of local anesthetics and one an overdose of morphine; it is not surprising that they experienced toxic complications. These three complications could have been avoided by using standard doses of drugs and by following the usual safety rules [1-3,23,24]
Peripheral nerve blocks, which are very useful for orthopedic procedures, represented only 38% of all regional blocks, and extremity nerve blocks only 16%. This low frequency of use is probably related to the rather poor reputation in pediatrics of nerve blocks, which are often supposed to be difficult and hazardous. This reputation was not confirmed in this study, as not even one adverse effect in 9396 procedures was reported after performance of peripheral nerve blocks and local anesthetics. This interesting finding should renew interest in these peripheral techniques and promote their more general use in pediatrics, provided the expected distribution of anesthesia is adequate and the anesthesiologist has sufficient experience in using the relevant block procedure.
Conclusion
This prospective study from May 1, 1993 to April 30, 1994 is representative of the pediatric anesthesia practice in France, Belgium, and one Italian institution. The first essential finding is that the use regional anesthesia is increasing in children, a development that seems logical as the overall morbidity of regional blocks is low, based on a group of 24,409 prospectively evaluated patients. The complications were rare, mostly minor, and did not result in sequelae or legal problems. Half of these complications could have been avoided by using appropriate equipment, a point still not sufficiently understood despite many publications stressing its importance.
Although several case reports of complications have overemphasized the negative side of pediatric regional block procedures, this survey demonstrates the safety of regional anesthesia in pediatric patients of all ages, including premature babies, based on data prospectively collected in a large number of children in several countries at the same time. Morbidity of peripheral nerve blocks and local anesthesia was extremely low (zero complications reported in this study): this should encourage anesthesiologists to use peripheral techniques as often as possible and when appropriate instead of using a central block including caudal anesthesia. The overall morbidity of central blocks was low but there were complications. Although none were detrimental to the patient or resulted in legal problems, half of these could have been avoided, whereas the other half were unavoidable. As a result, central blocks, including caudals, cannot be considered minor procedures: their indications must be weighed against the severity of the surgery and the physical status of the patient, not against the experience of the anesthesiologist with a particular technique. With simple guidelines for use, regional anesthesia should be considered a valuable and safe tool for providing high quality analgesia in pediatric patients.
The authors wish to thank Dr. Rita Khandwala for extensive revision of the manuscript and Ms. Harriet Felscher for editorial assistance. They also want to thank all the ADARPEF members who contributed to this prospective study, namely those from the following medical institutions: Angers (Hotel-Dieu), Besancon (Hopital Saint-Jacques), Bordeaux (Hopital Pellegrin), Boulogne-Billancourt (CCBB), Bruxelles (Clinique Universitaire Saint-Luc), Bruxelles (CHU Reine Fabiola), Caen (CHU), Charleroi (CHU A. Gailly), Clermont-Ferrand (Hotel-Dieu), Florence (Hopital A. Meyer), Grenoble (Hopital A. Michallon), Le Havre (CHG), Le Kremlin-Bicetre (Hopital de Bicetre), Lille (Hopital Saint-Antoine), Limoges (CHU Dupuytren), Marseille (Hopital Saint-Joseph), Marseille (Hopital Nord), Marseille (Hopital de la Timone), Montignee/Liege (Clinique de l'Esperance), Montpellier (Hopital Lapeyronie), Nancy (Hopital d'Enfants), Nice (Hopital Lenval), Orleans (CHR), Paris (Hopital Necker-Enfants Malades), Paris (Hopital Saint-Vincent de Paul), Paris (Hopital Trousseau), Pau (CHG), Reims (Memorial American Hospital), Rennes (Hopital Pontchaillou), Roscoff (Centre Helio-Marin), Rouen (Hopital Charles Nicolle), Saint-Brieuc (CH La Beauchee), Saint-Etienne (CHR), Saint-Etienne (Hopital Nord), Saint-Germain en Laye (CHG), Strasbourg (Hopital Hautepierre), Toulouse (Hopital Purpan), and Tours (Hopital Pediatrique Clocheville).
