A Perinatal Database Management System

Obstetric services have long recognized the need for ongoing evaluation of their experiences. Manual »departmental statistics« systems sufficed, but with the advent of more sophisticated care, perinatal re-gionalization and increased research activity, the potential usefulness of computer technology became obvious. At Cleveland Metropolitan General Hospital, a laboratory computer based patient information file system was designed and implemented beginning in 1974. Over the succeeding six years, data have been collected and stored for all delivered pregnancies. There are now over 61,000 files for more than 20,000 consecutively delivered patients. The system provides over 40,000 clinical reports per year. However, the use of a file-oriented system has limited our ability to respond to specific research queries. The application of a relational database management system, INGRES, for perinatal information is reported here. Examples of its use for efficiently »answering questions« are presented, as are guidelines for the development and implementation of computer-based perinatal record systems.

Correctly identifying the macrosomic fetus: improving ultrasonography-based prediction

OBJECTIVE

Our goal was to improve the accuracy of estimating fetal weights among macrosomic fetuses with the traditional measurements of abdominal circumference, femur length, and head circumference.

STUDY DESIGN

We used 4831 cases without anomalies from an ultrasonography laboratory database with an estimated fetal weight obtained a maximum of 14 days before delivery. Abdominal circumference, femur length, and head circumference were each regressed on birth weight to obtain estimated fetal weight by abdominal circumference, femur length, and head circumference, respectively. We compared the individual variation for estimated fetal weight by abdominal circumference, femur length, and head circumference by calculating a within-subject standard deviation to quantify the level of disparity. We adjusted the estimated fetal weight to the date of delivery and for dependencies on maternal diabetes mellitus, weight, and height. We then weighted cases with birth weight >4500 g and diabetic cases with birth weight >4000 g 20-fold (weighted estimated fetal weight) for the purpose of creating a favorable bias for classifying these cases. The equation of Hadlock et al, with abdominal circumference, femur length, and head circumference, was applied as a benchmark estimated fetal weight.

RESULTS

Of the 4831 newborns, 308 (6.4%) had a birth weight >4000 g, and 56 (1.2%) had a birth weight >4500 g. There were 154 pregnancies complicated by diabetes mellitus; 26 (16.9%) of the resulting infants weighed >4000 g, and 5 (3.2%) weighed >4500 g. At 95% specificity, the weighted estimated fetal weight had a sensitivity of 85.7% at a cut point of 3912 g, compared with a sensitivity of 71.4% at 3604 g by use of the estimated fetal weight of Hadlock et al.

CONCLUSIONS

We were able to improve the accuracy of identifying the macrosomic fetus compared to reliance on the equation by Hadlock et al. A fetus was found to be at significantly increased risk for birth weight >4000 g when the estimated fetal weight based on abdominal circumference is larger than that based on either head circumference or femur length or when there is a large within-subject variance in estimated fetal weight based on abdominal circumference, femur length, and head circumference. We also found that there were significantly different groups of patients whose estimated fetal weights require different equations for better estimates. Even given ultrasonographic measurements, taking into account maternal height, weight, and presence of diabetes mellitus can improve macrosomia detection. Although these findings remain to be optimized and validated, the approach used here appears to yield better predictions than the current "one function fits all" approach.

Variability of adjustments to indices in determining patient risk in biochemical screening

It has long been appreciated that the measurement of biochemical parameters for prenatal screening for neural tube defects, and later aneuploidy, is not as simple as measuring hemoglobin or hematocrit. Early in the game, it was recognized that there are gestational age curves, and that since alpha-fetoprotein (AFP), for example, is a fetal product, its distribution varies as a function of maternal plasma volume, and therefore the weight of the mother. A number of different adjustment factors have been used for AFP and other parameters for years, with varying degrees of consistency and reliability. Here we review a number of adjustments that have been used, and try to give priority to those that have been most effective. Furthermore, laboratories and programs need to be cognizant that with newer parameters being added, the specifics of requirements will vary on a case-by-case parameter basis, and optimal screening can only be achieved with the appropriate adjustments.

Survey of maternal-fetal medicine subspecialists: professional activities, job setting, satisfaction, and trends over time

The purpose of this survey was to describe the job setting, professional activities, and satisfaction of maternal-fetal medicine (MFM) subspecialists and to compare these with previous surveys. A questionnaire was mailed to members of the Society of Perinatal Obstetricians (SPO). The results were compared to a similar survey in 1986. Of 1,352 members, 58% responded. In comparison to 1986, there was an increase in percentage of women (25%); a decrease in university hospital-based faculty (55%); and an increase in community hospital-based salaried physicians (23%) and private practice (15%). The distribution of professional time changes with an increase in clinical care, a decrease in teaching and research. The annual number of procedures increased for ultrasound and genetics and decreased for gynecologic procedures. Since 1986, there have been significant changes among MFM subspecialists in job setting, allocation of professional time, and number and types of procedures. Job satisfaction remains high.

Precise gaussian distribution functions of maternal serum alpha-fetoprotein and free beta-subunit of human chorionic gonadotropin for trisomy 21 screening: improved accuracy for patient counseling

OBJECTIVE

Gaussian equation curves are used to generate baseline curves against which a priori maternal age Down syndrome risks are adjusted to develop likelihood ratios for individual patients. We sought to evaluate the accuracy of these calculations, minimize the affects of outliers, and to make improvements.

STUDY DESIGN

Gaussian distribution functions were used to investigate the best model for alpha-fetoprotein and free beta-human chorionic gonadotropin multiples of the median with use of nonlinear regressions. Parameters from distribution functions can be used to compute a more precise likelihood ratio for the decision logic for trisomy 21. A total of 58,297 normal cases and 348 cases of trisomy 21 were computed.

RESULTS

Log normal distribution functions generated by nonlinear regression produced excellent but exaggerated goodness of fit R2 to the frequency distributions of the data. For normal cases values were as follows (in mean, SD, and R2, respectively): log alpha-fetoprotein -0.07199, 0.15681, and 0.9970; log beta-human chorionic gonadotropin -0.15203, 0.24284, and 0.9987. For trisomy 21 cases the values were (in mean, SD, and R2, respectively) for log alpha-fetoprotein -0.19303, 0.15802, and 0.9828 and for log beta-human chorionic gonadotropin 0.19996, 0.29760, and 0.9669. Distributions reconstructed with use of statistical means and SDs generated goodness of fit R2 from 0.585 to 0.914. Use of means and SDs derived from distribution functions increased the R2 to 0.855 and 0.999. The change in the model produces, at a 5% false-positive rate, a sensitivity of 57.18% (199/348). A 1 in 113 cutoff point risk is obtained and is tighter than the 1 in 251 without the distribution functions, as versus 1 in 270 by age calculations alone.

CONCLUSIONS

Our data suggest that (1) normality of log transforms of alpha-fetoprotein and normality of log transforms of beta-human chorionic gonadotropin are reasonable models, (2) distribution functions can minimize the effect of outliers, which produces more realistic risk estimates, and (3) the effect of distribution functions versus standard mean and SDs cannot automatically be extrapolated to other parameters, which must be tested individually.

MOMs (multiples of the median) and DADs (discriminant aneuploidy detection): improved specificity and cost-effectiveness of biochemical screening for aneuploidy with DADs

OBJECTIVE

Our purpose was to assess the efficacy of double- and triple-screening paradigms for Down syndrome and to develop a more logical, statistical approach to risk prediction that will decrease the cost of screening and allow the incorporation of new parameters appropriately weighted for their contribution.

STUDY DESIGN

Data from 24,504 patients who had biochemical screening for Down syndrome by single (alpha-fetoprotein), double (alpha-fetoprotein, beta-human chorionic gonadotropin), or triple screening (alpha-fetoprotein, beta-human chorionic gonadotropin, unconjugated estriol) who had complete outcome information were analyzed by (1) existing gaussian-based methods, (2) the Glasgow ratio method, and (3) a new statistical approach (i.e., directly adjusted data sets for discriminant aneuploidy detection [DADs])

RESULTS

By use of individual risk-based thresholds for "at risk" status, both double and triple screening performed far better than single screening, but the percentages of patients at risk varied widely. When the percentages at risk were held constant, the sensitivity of double and triple screenings was similar, suggesting that there are no benefits of using estriol as a third marker. For 25,000 patients the use of only alpha-fetoprotein and beta-human chorionic gonadotropin would save about $500,000, with no decrease in sensitivity. With the DADs approach a statistically sound model giving more stable estimates was developed that permits each factor to be analyzed for its own explained proportion of variance and allows each parameter to have different weighting. For this data set the same sensitivity was seen with, conservatively, a 1% reduction in the percentage of patients at risk, which would reduce by 250 the number of amniocenteses, at a further savings of about $400,000.

CONCLUSIONS

By use of existing methods, double screening is equally as effective as triple screening, so that the expense of estriol is unnecessary. The DADs approach, by allowing for variable weighting of parameters, lowers the at risk percentage and will permit a much more flexible approach as new parameters become available. Changing to DADs and eliminating estriol should achieve higher specificity for the same sensitivity and save, conservatively, about $900,000 in this series. Extrapolated nationally, if confirmed, the annual savings could approach $72,000,000.

A gestational-age-independent model of birth weight based on placental size

A placental weight (P) to birth weight (W) relationship, W = P1.5 x constant (A), was presented in 1966. However, based on histological and theoretical studies, placental efficiency increases proportionally to the square of the gestational age in weeks. The purpose of this study was to develop and validate a gestational-age-independent model of birth weight based on P2. Data were prospectively entered into a computerized database for liveborn, structurally normal neonates at a tertiary, urban center. 29,989 neonates had gestational ages based on menstrual dating, ultrasound and Ballard examination, 81.4% were African American. The 'best fit' of W = A x P1.5 had errors in predicting the mean birth weight by at least 10% from 24 to 33 weeks of gestation. In contrast, there was significant improvement (p < 0.05) when the model W = 0.0071 x P2 was used; errors of greater than 10% only occurred at 28 and 43 weeks of gestation. This model had similar accuracy when stratified according to gender or ethnicity. A gestational-age-independent model of mean birth weight can be based upon the square of placental weight. We hypothesize that this model may have utility in understanding disorders of intrauterine fetal growth.

Approaching the millennium: perinatal problems and software solutions

Strategic planning for rational development of perinatal computing capabilities for the year 2000 should be driven by anticipated trends in (1) the health care business, (2) computer technology and (3) medicine, as well as (4) the needs of perinatal practitioners. In the USA, health care is the fastest growing segment of the economy. This will produce increasing attention from hardware and software developers, and vendors, and will lead to a proliferation of computing platforms, operating systems and specific medical application software. Desktop computers, already capable of 20 million instructions per second (MIPS) with massive storage capacities, will continue to evolve and fall in price. Increasingly, perinatologists will develop software packages to facilitate patient care in their own environments. All of these trends will lead to severe fragmentation in medical computing. Simultaneously, however, the need for integrated institutional computer-based data access for quality assurance and fiscal and operations management will increase. Perinatal care will be more regionalized, complex and rigorous with new clinical trial- and effectiveness research-based interventions, as well as molecular diagnosis and therapy. To practice appropriately, clinicians will need to be familiar with computer capabilities. Having been exposed to computer-aided instruction (CAI) at the undergraduate and postgraduate levels, they will except on-line access to detailed and accurate patient information with linkage to laboratory, radiology and other medical databases, as well as to reference databases, such as Medlines and the Oxford Database of Perinatal Trials. Artificial intelligence (AI) software may support perinatal decision making; computerized professional and facility billing will be available.

The mammary stimulation test--a predictor of preterm delivery?

Clinical risk identification for preterm delivery, the major cause of perinatal morbidity and mortality, remains problematic. To test the hypothesis that a provocative test for uterine contractility at the beginning of the third trimester would reliably identify patients destined to be delivered before and at term, we designed the mammary stimulation test. Uterine contractions in response to nipple stimulation at a mean of 28.6 weeks' gestation, were evaluated in 94 gravid patients at risk for preterm delivery. The mammary stimulation test, the results of which were not used for clinical management, was positive in 50% of patients tested and had a sensitivity of 84% and a positive predictive value of 34% (chi 2 = 11.15, p less than 0.01). Ninety-four percent of patients predicted to be delivered at term actually were delivered at term (negative predictive value). Furthermore, no patients with a negative mammary stimulation test were delivered within 1 month of testing or were delivered of infants weighing less than 2000 gm. Discriminant analysis indicated that the mammary stimulation test provided information beyond that available from clinical risk factors alone. A cost analysis suggested that the use of the mammary stimulation test could reduce the cost of ambulatory uterine activity monitoring by nearly 50%. If these findings can be validated in additional samples, the mammary stimulation test may be useful in prematurity prevention programs.

Ultrasound measurement of the fetal head: computer versus operator

In obstetrical ultrasound practice, biparietal diameter, occipito-frontal diameter and head circumference are among the most commonly used fetal measurements. To minimize the limitations associated with manual measurement, we have undertaken an investigation with a focus on (1) the design of a personal computer-based system for automated measurements of biparietal diameter, occipito-frontal diameter and head circumference, and (2) integration of such a system (including measurements of abdomen and femur) into the routine obstetrical ultrasound examination. This report presents preliminary results of a comparison of computer-determined fetal head measurements with those obtained by an operator. Data were obtained from 75 consecutive singleton fetal ultrasound examinations free of any obvious structural anomalies. The computer obtained acceptable measurements of biparietal diameter, occipito-frontal diameter and head circumference from 74 images and failed on one image. There was a highly significant correlation between computer-determined measurements of biparietal diameter (r = 0.986), occipito-frontal diameter (r = 0.958) and head circumference (r = 0.972) and those obtained by the operators. The mean measurement difference (computer minus operator) was 1.87 +/- 1.94 mm for biparietal diameter, 2.82 +/- 4.13 mm for occipito-frontal diameter and -0.36 +/- 9.87 mm for head circumference. These differences were independent of the operator's identity, the instrument used and gestational age. The key finding of this study is that, with the use of inexpensive personal computer technology, it is possible to design and implement a system that can give fetal head measurements which correlate highly with manual determination by a skilled operator and which take a fraction of the time.

Survey of maternal-fetal medicine subspecialists: professional activities, job setting, and satisfaction

We analyzed the job setting, professional activities, and satisfaction of subspecialists in maternal-fetal medicine. A two-page questionnaire was mailed to all members of the Society of Perinatal Obstetricians. The results were analyzed by multiple regression and multivariate analysis. Of 603 questionnaires mailed, 496 (80%) were returned; 70% of all certified maternal-fetal medicine subspecialists responded. The job setting of respondents was university hospital-based faculty in 66%, community hospital-based salaried in 15%, private practice in 10%, and other in 9%. The average percent of professional time devoted to each area was as follows: clinical work, 57% of all professional time; research, 18%; administrative, 15%; and educational, 11%. Time distribution, volume of hysterectomies, and number of peer-review publications were the only significant differences among job settings. For university hospital-based faculty, the time distribution was 50% clinical, 23% research, 16% administrative, and 11% educational. For community hospital-based salaried physicians, it was 63% clinical, 10% research, 16% administrative, and 11% educational; for private practice, it was 84% clinical, 7% research, 5% administrative, and 6% educational. Types of activities of maternal-fetal medicine subspecialists included ultrasound (75% of respondents, mean 436 studies per year), gynecologic procedures (71% of subspecialists, volume of 8.7 hysterectomies per year), genetic procedures (83%; 125 procedures per year), research (87%), recent peer-review publications (80%), grants (40%), and medicolegal reviews (64%). Ninety-five percent had a medical-school faculty appointment.(ABSTRACT TRUNCATED AT 250 WORDS)

A prototype system for perinatal knowledge engineering using an artificial intelligence tool

Though several perinatal expert systems are extant, the use of artificial intelligence has, as yet, had minimal impact in medical computing. In this evaluation of the potential of AI techniques in the development of a computer based "Perinatal Consultant," a "top down" approach to the development of a perinatal knowledge base was taken, using as a source for such a knowledge base a 30-page manuscript of a chapter concerning high risk pregnancy. The UNIX utility "style" was used to parse sentences and obtain key words and phrases, both as part of a natural language interface and to identify key perinatal concepts. Compared with the "gold standard" of sentences containing key facts as chosen by the experts, a semiautomated method using a nonmedical speller to identify key words and phrases in context functioned with a sensitivity of 79%, i.e., approximately 8 in 10 key sentences were detected as the basis for PROLOG, rules and facts for the knowledge base. These encouraging results suggest that functional perinatal expert systems may well be expedited by using programming utilities in conjunction with AI tools and published literature.

From mainframe to micro: decentralization of perinatal epidemiology

To evaluate the feasibility of analyzing large perinatal datasets on smaller computers, the 1980 National Natality Survey was downloaded first to a minicomputer and then to a microcomputer. Operating system utilities were used to minimize programming and storage requirements. Accuracy was confirmed by comparing descriptive statistics with published values derived on mainframes. Based on a typical analytic problem, implementation on the microcomputer compared favorably to implementation on the minicomputer. Our results suggest that the microcomputer may be a practical alternative to the mainframe for perinatal epidemiological analysis.

Maximizing the efficiency of a study on alcohol-related birth defects by means of data collection and analytic strategy dissociation

Application of the terms "prospective" and "retrospective" to quasi/experimental research designs used in clinical investigations is often ambiguous; they often lack specification as to whether they apply to data acquisition, analytic strategy, or both. Data may be collected prospectively by preplanned protocol or retrospectively by chart review. In addition, the direction of analysis may be prospective--comparing differential outcomes for at-risk and not at-risk (exposed/not exposed) groups--or retrospective--comparing differential preceding risks for affected and nonaffected groups. Prospective data collection is advantageous in assuring the availability of crucial variables for an entire sample, but prospective analytic strategies may be inefficient, underestimating effect sizes, particularly when the outcome of interest is rare or the effect, even if large, is seen in only a small proportion of cases. In a prospective multilinear regression model of risks for lowered birthweight for gestational age, 2% of the variance was explained by alcohol variables after adjustment for confounding. In comparison, a similar multilinear regression of case-control sample, in which the frequency of the abnormal outcome is raised, showed 18% of the variance explained by alcohol variables. Data collection and analytic strategies may be dissociated. Although some statistical power may be lost because of reduced sample size, synthetic case-control analysis is efficient in examining small effects. It also has the advantage of decreasing the size of the database to be managed and the time required for analysis. Using this approach, it is feasible to perform virtually any analysis that can be done on a mainframe on medium-sized departmental computer system.

The association between cesarean birth and outcome in vertex presentation: relative importance of birth weight, Dubowitz scores, and delivery route

The effect of the obstetrician's choice for delivery route on infant outcomes was studied in 17,260 mothers in labor, with the fetus alive at the onset of labor, in vertex presentation, and without major congenital anomalies. At birth weight categories, stratified by 500 gm increments, there was no improvement in intrapartum and neonatal mortality or in neonatal morbidity associated with the cesarean birth route. The major dependent variables important in neonatal outcome were birth weight and neonatal maturity. The association between cesarean section delivery and outcome in vertex presentations is limited. This information documents the need for prospective randomized and controlled studies with respect to proposed changes in birth route for the very low-birth weight fetus.

What affects fetal pulmonary maturation during diabetic pregnancy?

Previous studies differ as to whether the fetus of the woman with diabetes mellitus has altered formation of lung surfactant. The factors responsible for these differences remain unclear. In this study, measures of blood glucose control, including birth weight percentile and the presence of factors potentially influencing fetal pulmonary maturation, such as diabetic class, maternal chronic hypertension, and preeclampsia, were compared with the amniotic fluid lecithin/sphingomyelin ratio, phosphatidylglycerol, and phosphatidylinositol within a group of 90 diabetic pregnancies. The factors were evaluated in combination with the techniques of canonical correlation and multiple regression analysis. Gestational age had the strongest effect in determining levels of amniotic fluid phospholipids, and hypertension was found to significantly accelerate the appearance of phosphatidylgycerol (p less than 0.05). The effect of hypertension was one third as important as that of gestational age. Neither diabetic blood glucose control, White classification, nor the remaining explanatory variables were found to play a significant role in determining the amniotic fluid phospholipid levels. This study suggests that, in the clinical management of diabetes, gestational duration remains the single most important determinant of amniotic fluid phospholipid levels.

The effect of delivery route on outcome in breech presentation

This study was undertaken to determine the effect of choice of birth route on infant outcome in fetal breech presentation. The study group excluded infants who were footling breeches and infants with major congenital anomalies. The mothers were in labor with the fetal heartbeat present at entry into the labor room. Outcome variables were intrapartum and neonatal deaths and neonatal neurological morbidity. Delivery route was not significantly associated with neonatal death in the 500 to 999 gm (p = 0.43) and 1,000 to 2,499 gm (p = 0.43) categories. Over 2,500 gm, there were no neonatal deaths. In similar manner, delivery route was not significantly associated with neonatal neurological morbidity. In both neonatal death and morbidity, birth weight and modified Dubowitz score explained the largest proportion of the variance for outcome in breech presentation.

Risks preceding increased primary cesarean birth rates

Recent increases in cesarean birth rates are of concern to obstetricians and to the public. Previous studies often focused on indications for abdominal delivery, described in such broad categories as repeat cesarean, breech presentation, dystocia, and fetal distress, without detailed information concerning medical/obstetric risk factors. In this study of 2744 consecutively delivered mothers who received antenatal care, the problem was approached differently by evaluating approximately 100 pregnancy risk factors recorded in a computerized uniform perinatal record system before making a decision for delivery. Excluding repeat cesareans, relative risks for primary cesarean birth were calculated for each risk factor. For the large majority of patients, the need for primary cesarean birth could not be predicted before the onset of labor. Approximately 70% of primary cesareans could be accounted for on the basis of single pregnancy risks--antepartum problems, 11%; abnormal fetal presentation, 30%; abnormal labor, 14%, and fetal distress, 15%. This leaves 30% of primary cesareans without a single major preceding risk. Along with the facts that 80% of primary cesareans in this series were performed for normal-size infants and that dystocia appears to account for a 3.4-fold greater proportion of cesareans in the national experience than in the current study, this suggests that critical evaluation and study of obstetric management of dystocia may be most fruitful in responding to the question of rising cesarean birth rates.

Regular and irregular human fetal respiratory movement

Fetal respiratory movements (FRM) were studied using abdominal strain gauges (tocodynamometers). The patterns of the FRM were evaluated during both active and quiet fetal time periods, which were determined by the fetal heart rate (FHR) and fetal body movement (FM). The FRM were classified into Regular and Irregular patterns based on neonatal respiratory criteria for sleep-state studies in the term infant. Evaluation of the breath-to-breath intervals (BBI) showed statistically significant respiratory differences during active and quiet fetal time periods. Irregular fetal respiratory movement patterns were noted during fetal active periods. It would appear that the correlation of regular fetal respiratory movement with fetal quiet periods in the term fetus adds additional evidence that a quiet sleep state may exist in the term fetus.

Use of computers in the labor and delivery suite: an overview

Some simple computers for fetal monitoring are already commercially available. These systems are designed to alarm when abnormalities reflected by the fetal heart are detected during labor. Unfortunately, the problems of fetal heart rate/intrauterine pressure data analysis have been oversimplified. An "ideal" obstetric computer-based data management system would begin, as the clinician should, with the assessment of risk from the patient's past history, pregnancy course, and labor progress as a basis for analyzing the condition of the fetus during labor. Systems available now are of limited scope, expensive, and of unproved clinical value. Premature attempts at clinical computerization will accomplish little, except to give computers a bad name. Let the buyer beware!

Clinical estimation of gestational age: rules for avoiding preterm delivery

Reliable knowledge of the duration of pregnancy prior to birth is often of crucial importance in making obstetric care decisions. Laboratory methods for estimating fetal maturity have received considerable attention, but the usefulness of historical information has only rarely been addressed. In order to examine the value of clinical estimators of fetal gestational age (GA) in 690 pregnancies, the correlations of menstrual history (LMP), first unamplified audible fetal heart tones (FFH), and quickening (Q), with GA, based on the modified Dubowitz examination at birth, were examined. Evaluation of each of the data sets used alone reveals that in order to be 90% certain that an infant will be mature at delivery (greater than or equal to 38 weeks), a reliable LMP must have been noted for 42 weeks prior to birth, the FFH heard for 21 weeks, and Q felt for 25 weeks. These findings suggest that carefully obtained historical and physical examination information remains a cornerstone of appropriate obstetric care.

Clinical application of high-risk scoring on an obstetric service

Obstetric risk scoring is a formalized way of recognizing, documenting, and cumulating antepartum and intrapartum factors to predict later complications for mother, fetus, and infant. If simple, practical, and reliable, risk scoring can be clinically useful in determining appropriate levels of care. In this prospective study, antepartum and intrapartum risk scales were integrated into the clinical record, and the relationship of risk scores to outcome was evaluated for 1,275 consecutively delivered gravid women. The forms could be simply and quickly filled out by the staff. Increased risk on both scales was significantly related to lowered one- and five-minute Apgar scores. The perinatal mortality rate increased from 0 to 93.4 per thousand from the lowest to the highest risk group. More than 80% of all perinatal deaths occurred in the one quarter of patients in the highest risk group. These results suggest that this risk scoring system can be used effectively in a clinical setting to identify patients at increased risk for neonatal depression and perinatal death.

Computer interpreted fetal monitoring data. Discriminant analysis or perinatal data as a model for prediction of neurologic status at one year of age

Intrapartum fetal monitoring may be useful in discriminating short-term and long-term infant outcomes. In order to develop a model for the classification of infants for measures of outcome and to evaluate the discriminatory value of selected perinatal data, 61 data files, which included computer interpreted fetal electroencephalogram and fetal heart rate, one and five minute Apoar scores, results of neonatal neurologic examination, and results of neurologic examination at one year of age, were submitted to computerized discriminant function analysis. For the classification of infants for neurologic status at one year, using FEEG patterns alone, 64% of the 42 normal infnats and 63% of the 19 abnormal infants were correctly classified. Using intrapartum FEEG and FHR patterns simultaneously, 69% of the 42 normal infants and 63% of the 19 abnormal infants were correctly classified. Combining intrapartum data with postpartum data, including one minute and five minute Apgar scores and neonatal neurologic examinations, 81% of the 42 normal infants and 79% of the 19 abnormal infants could be correctly classified. These results suggest that brain damage may already be present before birth, during labor, and that the risk of abnormal neurologic outcome at one year can be detected for a large percentage of infants using computer interpreted monitoring data during labor.

Computer interpreted fetal electroencephalogram: sharp wave detection and classification of infants for one year neurological outcome

The presence of visually discernible sharp waves (SWs) in the fetal electroencephalogram (FEEG) has been found to be associated with abnormal neurological infant outcome, but no method of programmed SW detection for FEEG was available. In order to develop an algorithm for SW detection, the first and second derivatives for visually identified SWs and non-SWs were examined and five random variables chosen for discriminant function analysis (DFA). The resulting equation, incorporated into program logic along with logic for artifact rejection, produced classifications from 85% to 89% consistent with visual identifications, suggesting that the number of SWs/epoch (NSW) corresponds with visually identified SWs. In addition, in 61 cases using a threshold for NSW derived by DFA, computer recognized SWs were found to be significantly related to the overall visual interpretation of the tracings (P less than 0.005). Finally, NSW alone produced correct classification of 65.5% of infants for 1 year neurological outcome. The overall consistency was increased to as high as 80% using additional FEEG and neonatal data. These findings imply that some forms of brain damage are present before birth and can be detected during labor using FEEG.

Normal and abnormal labor progress: I. A quantitative assessment and survey of the literature

Quantitative labor data have been reported piecemeal in the literature; there is no previous unified report of the frequencies of each of the dysfunctional labor patterns (DLPs). In order to describe the labor process more clearly a labor diagnostic computer program was used to prospectively study an unselected series of 3,682 nearly consecutive labors from one institution. Cervicometric data for nulliparous and multiparous labor were very similar to the results of others. The absolute frequency with which each DLP was found to occur is compared with results in the literature. Protracted active phase dilatation was the most frequently occurring DLP and prolonged deceleration phase the least frequently occurring one. The relative frequencies of the DLPs, defined as the percentage of all DLPs that occurred, remained reasonably constant, regardless of parity or the absolute frequencies of the DLPs in a group of patients.

"Prediction" of the one-minute Apgar score from fetal heart rate data

The value of any fetal monitoring technic is in its ability to predict infant outcome. In the present study, the ability of fetal heart rate (FHR) monitoring data to "predict" a measure of short-term infant outcome, the 1-minute Apgar score, was evaluated using univariate and multivariate statistical analyses. Of 61 monitored high-risk infants, 46 had high (7 to 10) and 15 had low (1 to 6) 1-minute Apgar scores. Computer analysis of FHR/intrauterine pressure (IUP) data for these 61 infants revealed that the infants with low Apgar scores had more than the expected number of late decelerations (LD). Using a threshold of ten LD and univariate analysis, 74% of the infants could be properly classified for high or low Apgar scores, but 60% of the infants with low Apgar scores were not identified. Using discriminant function (multivariate) analysis for the numbers of LD and uterine contractions, 47% of the depressed infants were appropriately identified and simple risk scoring equations were devised. Using additional observation vectors, including the number of accelerations and early decelerations, 67% of the depressed infants could be identified. The results of this study suggest that using multiple observation vectors improves the predictive capacity and, thus, the value of fetal monitoring data. Clinical experience suggests that the value of monitoring data can be further enhanced by simultaneous evaluation of other observation vectors from additional perinatal data sets using the technics of this study.

Computer interpreted fetal electroencephalogram. II. Patterns in infants who were neurologically abnormal at 1 year of age

A computer program for pattern recognition of fetal electroencephalogram has been used to analyze the records of nine fetuses, known to be neurologically abnormal at 1 year of age. In 4,913 10 second epochs of adequate FEEG, Low Voltage Irregular (LVI) accounted for 17.8 per cent, Mixed activity (MIX) for 30.5 per cent, High Voltage Slow (HVS) for 18.1 per cent, and Trace Alternant (T/A) for 33.2 per cent of the epochs. The numbers of observed FEEG patterns in these abnormal cases appear to be significantly different from those in 11 normal cases (p less than 0.001)3. Specifically, the relative frequency of LVI was found to be increased in the abnormal cases (p congruent to 0.05). Moreover, LVI was significantly associated with visually interpreted prolonged voltage suppression (p less than 0.025) and lowered one-minute Apgar score (p congruent to 0.025). Using discriminant function analysis for LVI, MIX, HVS, and T/A patterns from FEEG recorded during labor, 10 of 11 infants were correctly classified as being neurologically normal at one year of age and 6 of 9 infants were correctly classified as being neurologically abnormal at 1 year of age. These studies confirm previous associations based on visual analysis of FEEG and suggest that the relative frequencies of FEEG patterns may be useful in the prediction of neurologic outcome 1 year later.

Computer interpreted fetal electroencephalogram. I. Relative frequency of patterns

Fetal electroencephalography (FEEG) is a clinical research technique for monitoring the electrical activity of the fetal brain during labor. Because of the massive volume of data, it has not previously been practicable to quantify the frequency of occurrence of the various patterns by visual analysis. With the use of a computer program, which has been shown to be consistent with visual interpretations, records for 11 fetuses, known to be neurologically normal at 1 year of age, were analyzed. Adequate FEEG had been recorded 73.6 per cent of the time during which the fetuses were monitored. In 10,511 10 second epochs of adequate FEEG, the Mixed pattern was found to be dominant, with a relative frequency of 41.2 per cent. The relative frequencies of Trace Alternant and High Voltage Slow activity were 32.3 per cent and 21.5 per cent, respectively. Low Voltage Irregular, Voltage Depression, and Isoelectricity occurred infrequently, accounting for approximately 4.6 per cent of the epochs. The relative frequencies of the various patterns were stable from the onset of monitoring of FEEG through neonatal EEG obtained in the delivery room. These data will form the basis for subsequent studies of FEEG changes associated with various endogenous and exogenous factors.

An interactive computer program for studying fetal electroencephalograms

Fetal electroencephalogram (FEEG), recorded during labor, produces very large volumes of data for visual interpretation. An established terminology, developed for the interpretation of neonatal electroencephalogram, has been found to be useful for visual pattern recognition of FEEG. A program, which identifies FEEG patterns within ten second epochs and provides direct comparison between visual and programmed analysis, has been developed using an interactive computer system. This program provides 85-90 percent consistency with visual interpretation.

Fetal electroencephalography. Relationship to neonatal and one-year developmental neurological examinations in high-risk infants

Sharp wave activity in fetal electroencephalograms was found to have a significant relation to neurological findings at 1 year of age but not to neurological findings obtained during the neonatal period. Neonatal neurological findings were significantly related to those at 1 year of age.