Lisa J Jones, Paul D Craven, Anil Lakkundi, Jann P Foster, Nadia Badawi

With improvements in neonatal intensive care, more preterm infants are surviving the neonatal period and presenting for surgery in early infancy. Inguinal hernia is the most common condition requiring early surgery, appearing in 38% of infants whose birth weight is between 751 grams and 1000 grams. Approximately 20% to 30% of otherwise healthy preterm infants having general anaesthesia for inguinal hernia surgery at a postmature age have at least one apnoeic episode within the postoperative period. Research studies have failed to adequately distinguish the effects of apnoeic episodes from other complications of extreme preterm gestation on the risk of brain injury, or to investigate the potential impact of postoperative apnoea upon longer term neurodevelopment. In addition to episodes of apnoea, there are concerns that anaesthetic and sedative agents may have a direct toxic effect on the developing brain of preterm infants even after reaching postmature age. It is proposed that regional anaesthesia may reduce the risk of postoperative apnoea, avoid the risk of anaesthetic-related neurotoxicity and improve neurodevelopmental outcomes in preterm infants requiring surgery for inguinal hernia at a postmature age. To determine if regional anaesthesia reduces postoperative apnoea, bradycardia, the use of assisted ventilation, and neurological impairment, in comparison to general anaesthesia, in preterm infants undergoing inguinal herniorrhaphy at a postmature age. The following databases and resources were searched: the Cochrane Central Register of Controlled Trials (CENTRAL, The Cochrane Library, 2015, Issue 2), MEDLINE (December 2002 to 25 February 2015), EMBASE (December 2002 to 25 February 2015), controlled-trials.com and clinicaltrials.gov, reference lists of published trials and abstracts published in Pediatric Research and Pediatric Anesthesia. Randomised and quasi-randomised controlled trials of regional (spinal, epidural, caudal) versus general anaesthesia, or combined regional and general anaesthesia, in former preterm infants undergoing inguinal herniorrhaphy in early infancy. At least two of three review authors (LJ, JF, AL) independently extracted data and performed analyses. Authors were contacted to obtain missing data. The methodological quality of each study was assessed according to the criteria of the Cochrane Neonatal Review Group. Data were analysed using Review Manager 5. Meta-analyses were performed with calculation of risk ratios (RR) and risk difference (RD), along with their 95% confidence intervals (CI) where appropriate. Seven small trials comparing spinal with general anaesthesia in the repair of inguinal hernia were identified. Two trial reports are listed as 'Studies awaiting classification' due to insufficient information on which to base an eligibility assessment. There was no statistically significant difference in the risk of postoperative apnoea/bradycardia (typical RR 0.72, 95% CI 0.48 to 1.06; 4 studies, 138 infants), postoperative oxygen desaturation (typical RR 0.82, 95% CI 0.61 to 1.11; 2 studies, 48 infants), the use of postoperative analgesics (RR 0.42, 95% CI 0.15 to 1.18; 1 study, 44 infants), or postoperative respiratory support (typical RR 0.09, 95% CI 0.01 to1.64; 3 studies, 98 infants) between infants receiving spinal or general anaesthesia. When infants who had received preoperative sedatives were excluded, the meta-analysis supported a reduction in the risk of postoperative apnoea in the spinal anaesthesia group (typical RR 0.53, 95% CI 0.34 to 0.82; 4 studies, 129 infants). Infants with no history of apnoea in the preoperative period and receiving spinal anaesthesia (including a subset of infants who had received sedatives) had a reduced risk of postoperative apnoea and this reached statistical significance (typical RR 0.34, 95% CI 0.14 to 0.81; 4 studies, 134 infants). Infants receiving spinal rather than general anaesthesia had a statistically significant increased risk of anaesthetic agent failure (typical RR 7.83, 95% CI 1.51 to 40.58; 3 studies, 92 infants). Infants randomised to receive spinal anaesthesia had an increased risk of anaesthetic placement failure of borderline statistical significance (typical RR 7.38, 95% CI 0.98 to 55.52; typical RD 0.15, 95% CI 0.03 to 0.27; 3 studies, 90 infants). There is moderate-quality evidence to suggest that the administration of spinal in preference to general anaesthesia without pre- or intraoperative sedative administration may reduce the risk of postoperative apnoea by up to 47% in preterm infants undergoing inguinal herniorrhaphy at a postmature age. For every four infants treated with spinal anaesthesia, one infant may be prevented from having an episode of postoperative apnoea (NNTB=4). In those infants without preoperative apnoea, there is low-quality evidence that spinal rather than general anaesthesia may reduce the risk of preoperative apnoea by up to 66%. There was no difference in the effect of spinal compared with general anaesthesia on the overall incidence of postoperative apnoea, bradycardia, oxygen desaturation, need for postoperative analgesics or respiratory support. Limitations on these results included varying use of sedative agents, or different anaesthetic agents, or combinations of these factors, in addition to trial quality aspects such as allocation concealment and inadequate blinding of intervention and outcome assessment. The meta-analyses may have inadequate power to detect a difference between groups for some outcomes, with estimates of effect based on a total population of fewer than 140 infants.The effect of newer, rapidly acting, quickly metabolised general anaesthetic agents on safety with regard to the risk of postoperative apnoea and neurotoxic exposure has not so far been established in randomised trials. There is potential for harm from postoperative apnoea and direct brain toxicity from general anaesthetic agents superimposed upon pre-existing altered brain development in infants born at very to extreme preterm gestation. This highlights the clear need for the examination of neurodevelopmental outcomes in the context of large randomised controlled trials of general, compared with spinal, anaesthesia, in former preterm infants undergoing surgery for inguinal hernia.There is a particular need to examine the impact of the choice of spinal over general anaesthesia on respiratory and neurological outcomes in high-risk infant subgroups with severe respiratory disease and previous brain injury.