Risk factors for apnea after infant inguinal hernia repair

SIR—We read with interest the letter by Kim et al. describing their experience with apnea and hernia repair in infants (1). We have also performed a retrospective audit, and given the paucity of published data, wish to present our findings which reinforce many of Kim et al.’s conclusions. The aim of our audit was to determine risk factors for spinal anesthesia failure and postoperative apnea. After approval from our Human Research Ethics Office we searched our theatre database to identify infants aged <1 year of age who had hernia repair between 1 July, 2006 and 31 December, 2006. We identified 129 infants. Records were missing for two infants. Of the rest, 91 had planned general anesthesia (GA), 29 had successful spinal anesthesia with no supplemental sedation apart from oral glucose and seven had unsatisfactory spinal anesthesia requiring sedation or conversion to GA. Of the seven in the unsatisfactory spinal anesthesia group, two spinals were abandoned because of total failure while five required only brief supplemental sedation. The gestational age at birth, postmenstrual age (PMA) and weight for each anesthesia group are shown in Table 1. Spinals tended to be attempted in children with lower weight, PMA and gestational age at birth, but there was considerable overlap between groups. In two cases, spinal failures were because of inexperience, in one case because of a bloody tap and for four babies there was no obvious reason for failure. Analyzing the 36 babies where spinal anesthesia was attempted and taking failed spinal as the outcome, a logistic regression found no association between failure and PMA, gestation at birth, weight and dose of local anesthetic. However, the numbers are small, so weak associations cannot be ruled out. Apnea was defined as any apnea observed and recorded in the patient record, irrespective of duration or heart rate. Apnea was defined as early if occurring within the first hour, and late if occurring in hospital after 1 h. Early apnea was not recorded in any child who had a successful spinal anesthetic, however similar to Kim et al. we found that in children where spinal anesthesia was unsuccessful, early apnea was particularly common (Table 2). One child had a long apnea at 21 h after an uncomplicated general anesthetic resulting in a brief unplanned admission to ICU. This child was 28 weeks gestation at birth, 41 weeks PMA and weighed 2.7 kg. We used logistic regression analysis to define risk factors for early and late significant apnea. For the regression, intention-to-treat (GA vs intention for spinal which includes failed spinal and successful spinal) and as-per-protocol analyses (GA and unsuccessful spinal vs successful spinal) were both performed. For the logistic regression analysis, significant apnea was defined as any baby having an apnea apart from those who had a single untreated apnea that required no intervention and was not associated with desaturation. By logistic regression with an intention-to-treat analysis, there was no evidence for an association between significant early apnea and PMA, gestation at birth, weight or intention for spinal anesthesia. Similarly with an as-perprotocol analysis there was no evidence for an association between significant early apnea and gestation at birth, PMA or weight; however having a successful spinal could not be entered into this particular regression as none of those with successful spinal had any early apnea. In summary for early apnea it is difficult to predict which baby will have an early apnea, but if a spinal was given and successful, then the risk can be assumed to be low and if a spinal was given and failed, the risk for early apnea is high. For late apnea, there was weak evidence of an association between significant apnea and lower PMA (P = 0.08) but no association between significant late apnea and intention-to-give a spinal, gestation at birth or weight. With the as-per-protocol analysis there was also an association between significant late apnea and lower PMA but no association between significant late apnea and gestation at birth, weight or successful spinal. In summary, late apnea is more strongly associated with low PMA than type of anesthesia, gestation at birth or weight.