Predictive Value of Monitoring Parameters in Fetal Surgery By Franqois I. Luks, Brian D. Johnson, Konstantinos Papadakis, Mohamed Traore, and George J. Piasecki Providence, Rhode Island Background/Purpose: The choice of monitoring parameters in fetal surgery has thus far been based on feasibility rather than on predictability. Ideally, monitoring should be noninva- sive, have a rapid response time and high sensitivity, and be applicable to open and endoscopic techniques. Herein, the authors studied the response of several parameters to stan- dardized episodes of fetal ischemia and stress. Methods: Eight time-dated fetal lambs (110 days, term, 145 days) were used. Under general anesthesia, a balloon oc- eluder was placed around the umbilical cord. Pulse oximetry (POX + heart rate, HR), electrocardiography (ECG), direct oximetry (DOx), and blood pressure (BP) were recorded continuously. After stabilization, the umbilical cord was com- pletely occluded for 5 seconds, then released. False-negative recordings were defined as failure of a parameter to respond to umbilical occlusion; false-positive episodes were defined as 10% change in value over 510 seconds during stabilization (baseline) period. Resu/ts:The fetuses were monitored for an aggregate of 358 minutes. Baseline DOx was 64% ? 5%, POX, 66% ? 16%; HR, 141 2 18 beats per minute (bpm); systolic BP (SBP), 51 t 3 torr; and diastolic BP (DBP), 38 -C 2 torr. During umbilical occlusion (n = 151, SBP increased to 56 -C 3 torr and DBP to 43 t 2 torr at 0.5 seconds, then returned to baseline at 8.0 seconds. A decrease was seen in DOx (start at 3.5s, maximum n 9.9 + 1.5% at 10.5 seconds) and POX (start at 4.2 seconds, maximum n 7.3 2 2.4% at 20.5 seconds). Heart rate showed <IO% decrease (start at 8.5 seconds, nadir 131 -C 14 bpm at 19.5 seconds). No ECG changes were noted. Sensitivity was 100% for DOx, POX, and BP, but only 14% for HR; specificity was 97% for DOx and 88% for POX; positive predictive value was 58% for DOx and 37% for POX; negative predictive value was 100% for DOx and POX. Conclusions: Direct intravascular oximetry and blood pres- sure provide a prompt and reliable response to acute fetal stress, but are too invasive for routine use. Bradycardia is an insensitive and late sign of fetal distress. Pulse oximetry has a rapid response time (~5 seconds), high sensitivity, and negative predictivevalue. In addition, its application is nonin- vasive and has proven to be feasible in open and endoscopic fetal surgical procedures. It therefore appears to be the monitoring parameter of choice for fetal surgery. J Pediatr Surg 33:1297-1301. Copyright o 1998 by W.B. Saunders Company INDEX WORDS: Fetal surgery, monitoring, oximetry, oxygen saturation, fetal distress. F ETAL SURGERY was introduced as a semiexperi- mental treatment modality more than a decade ago. l It has gained momentum in recent years, thanks to novel alternatives in the treatment of certain conditions such as diaphragmatic hemia2-4 and the development of newer, less invasive access techniques to the uterus.5-g As with operations on any other patient, surgery on the fetus requires adequate intraoperative monitoring. Until now, fetal monitoring during in utero surgery has been based on feasibility, rather than on accuracy and predic- tive value. Electrocardiography (ECG), monitoring of heart rate, temperature, and pulse oximetry have all been tried with varying results. l,i”-i3 Some, such as tempera- ture and amniotic pressure, monitor long-term homeosta- sis during fetal intervention. However, none of these parameters have been studied for their capability to rapidly and accurately detect fetal distress. The ideal monitoring parameter should (1) have a short lag time (rapid response), (2) have a high sensitivity (ie, detect all episodes of stress), (3) have a high negative predictive value (ie, no change in parameter means that the fetus is doing well), (4) be noninvasive and, particu- larly in view of newer uterine access techniques, (5) be Journal ofpediatric Surgery, Vol33, No 8 (August), 1998: pp 1297-1301 applicable to both open and endoscopic fetal surgery. Added advantages are a high specificity and positive predictive value (ie, a minimum of false alarms), al- though these characteristics are clearly secondary to a high sensitivity. To determine the value of several known fetal monitoring parameters, a reproducible method of reversible fetal stress and hypoxia was developed in the fetal lamb. MATERIALS AND METHODS Eight time-dated fetal lambs (110 days, term, 145 days) were used. After placing the ewe under general halothane anesthesia, a laparotomy exposed the gravid uterus. A limited hysterotomy was made, and the umbilical cord was encircled with a balloon occluder (30~mL. silicone balloon). The following monitoring devices were placed on or in the fetus: ECG electrodes (two on the fetal chest, one [reference] electrode on the myometrium); waterproof peripheral pulse oximetry probe From the Division of Pediatric Surgery, Brown University School of Medicine, Providence, RI. Address reprint requests to Fraqois I. Luks, MD, Dwision of Pediatric Surgery, Hasbro Children’s Hospital, 2, Dudley St, Suite 180, Providence, RI 02905. Copyright o 1998 by WB. Saunders Company 0022-3468/98/3308-0026$03.00/O 1297