History of Fetal Monitoring

The status of the fetus during labor has been monitored for centuries. In fact listening to the fetal heart rate through a stethoscope (auscultation) has been a part of labor management for more than 100 years. Fetal bradycardia (abnormally slow heart rate) and meconium staining of amniotic fluid were recognized as indicators of fetal distress in the late nineteenth century.¹ Attempts to record the fetal heart rate (FHR) began as early as 1891; however, current electronic methods of fetal monitoring have been developed since 1960.

In 1957 Hon first reported the successful recording of the fetal electrocardiogram (EKG) from the mother's abdomen.² In 1959 he observed profound bradycardia in a dying fetus and noted that manipulation of the prolapsed cord caused cardiac irregularities and bradycardia.³ In 1960 Hon described a clinical FHR monitor, although he recognized the need for a better method of recording.⁴ According to Hehre, Hon conceived the idea of passing an electrode through the cervix and clipping it to the fetus's head to record the EKG; he made a prototype of the device in his home workshop and appeared for the initial testing of it in the delivery room at 3 a.m.⁵ The electrode was successful and became the basis for present EFM devices; a major improvement, the spiral electrode, was developed by Hon in 1972.⁶

By 1969, EFM by ultrasound and by direct EKG monitoring had begun to diffuse rapidly. Hon and Caldeyro-Barcia were the most active investigators, describing a variety of changes in the FHR and correlating them with clinical status. Working independently, these two men devised different schemes for interpreting FHR patterns. Because of the resulting confusion in the obstetrical community, they agreed on a standard terminology in the 1970s.⁷

In 1961 Saling reported a technique for sampling fetal scalp-blood.⁸ In this technique a scalpel is passed through the cervical os, and a small wound is made in the baby's scalp. A sample of blood is then collected in a capillary tube and analyzed for acid-base parameters, pH being the most widely accepted measure of fetal well-being. This technique has gradually come to be considered an integral part of EFM and is often used to validate a diagnosis made by FHR monitoring.

At first, EFM was used largely for high-risk pregnancies, and some physicians continue to advocate its use primarily for that group. An increasing number of obstetricians, however, favor EFM during all labors, and many institutions in this country and elsewhere use it routinely. By 1972 an estimated 1,000 EFM systems were in use in the United States, and a survey in 1976 reported that 278 of 279 obstetrics programs with residencies in obstetrics used EFM.⁹ Surveys of physicians reported in 1970 and 1976 found a high degree of acceptance of the procedure, and the 1976 survey showed that 77 percent of the physicians surveyed believed that EFM should be used in all labors.^{10, 11}

Today, EFM is done externally (using ultrasound), internally (by attaching electrodes directly to the fetus to monitor its EKG), or sequentially (using both techniques before and after the amniotic sac ruptures). For internal monitoring, a catheter is used to monitor uterine contractions.

Despite the widespread application of EFM and an extensive literature on the interpretation of results, there is surprisingly little data on its pattern of use in the community. Several reports have described patterns of use at individual institutions, however. Initial reports from several university or large teaching programs demonstrated that clinicians attempted to use EFM for all labors within two years of purchasing a monitor. At these institutions, the percentage of pregnancies for which EFM was used ranged from 86 to 100.¹² Later reports indicated less widespread monitoring at community-based institutions.

Few population-based studies have been performed. One based on 1978 birth records from upstate New York, however, found that 47 percent of live births had some form of electronic monitoring, with external EFM predominating.¹³ Data from the 1980 National Natality Survey were almost identical: EFM was used in 47.7 percent of live births, and EFM predominated.¹⁴ By 1986, 75 percent of live births in New York State were monitored electronically, most of them externally (D. Mayack, personal communication, 1988). The National Natality Survey will include detailed questions about EFM when it is reported in 1990 (P. Placek, personal communication, 1988).

Advocates of EFM promised that its use would reduce the rate of intrapartum stillbirth, neonatal illness and death, and developmental disabilities. Early, uncontrolled observations showed a reduction in both perinatal mortality and low Apgar scores in association with the increased use of EFM during labor. Moreover, labors monitored with EFM had superior outcomes compared with other, less complicated labors without EFM in the same institution. Nonetheless, widespread introduction of this technology prompted national controversy in the 1970s, controversy that was heightened after four randomized clinical trials indicated little or no benefit with EFM.¹⁵ Reports of five additional clinical trials have been published—all but one with negative results—yet established policies for intrapartum surveillance with EFM have not been altered as a result. ^{16, 17} Although the controversy surrounding EFM has diminished substantially during the 1980s, the proper role of EFM in intrapartum care has not been settled.

Efficacy and Safety of Electronic Fetal Monitoring

Technology assessment is the systematic study of the possible effects on society of new, extended, or modified technology, with special emphasis on impacts that are unintended, indirect, and delayed.¹⁸ The assessment's purpose is to provide decision makers with information on policy alternatives, such as allocating research and development funds, formulating regulations, or developing new legislation. The ultimate goal of technology assessment, however, is to improve outcome.

The congressional Office of Technology Assessment took the lead in technology assessment in the United States in the 1970s, and now interest and activities are widespread in government, universities, and the private sector. At the Institute of Medicine, the Council on Health Care Technology addresses central issues in technology assessment through panels on information, methods, and evaluation.

The basic task of technology assessment is to document efficacy and safety. The key tool in the evolution of efficacy is the randomized clinical trial (RCT). Primary clinical evaluations can be ranked according to their freedom from bias: RCTs appear at the top, followed by nonrandomized controlled studies, series of patients without controls, and personal recollection unaided by systematic recordkeeping. Yet with the notable exception of drug trials, RCTs are rarely conducted before a technology has been diffused. Because of their expense, limited generalizability, and difficulty of implementation, RCTs rarely provide sufficient evidence to encourage or limit the diffusion of a new technology or to necessitate the withdrawal of a widely used technology. A thorough technology assessment is expensive and time-consuming and is simply not practical for all medical technologies. Even when well conducted, the findings may be overridden by social, economic, or political considerations.

None of the nine RCTs noted in the last section demonstrated a statistically significant decrease with the use of EFM in the rates of perinatal death, intrapartum stillbirth, neonatal death, 1-minute Apgar score of < 7, 1-minute Apgar score of < 4, or neonatal intensive care admissions (Table 1). The first Melbourne trial and the second Denver trial showed a decrease in neonatal seizures associated with EFM and the use of fetal scalp-blood sampling.^{19, 20} Whereas the decrease was not statistically significant, this observation confirmed the finding in the Dublin study.²¹ The first Melbourne study suggested a significant decrease associated with EFM in the rate of admission to neonatal intensive care units, associated with EFM, a finding not confirmed elsewhere. All of the trials showed an increased rate of cesarean delivery in the EFM group. The results for total operative deliveries were mixed: increases in the rates of operative deliveries (that is, abdominal and forceps deliveries) were reported in the EFM group in both Melbourne trials, the second Denver trial, and both the Copenhagen and Dublin studies; decreases were reported in the EFM group in the first Denver and the Sheffield trials.^22-28

TABLE 1

Results of Nine Clinical Trials Assessing the Efficacy of Routine Electronic Fetal Monitoring During Labor.

The pooled data^* showed slight but not significant increases in the EFM group in the number of low Apgar scores, the number of admissions to neonatal intensive care units, and the perinatal death rate (Table 2). The EFM group did, however, show a statistically significant decrease in the number of neonatal seizures. It also had statistically significant increases in the rate of both abdominal and total operative deliveries.

TABLE 2

Pooled Data from Eight Randomized, Controlled Trials of Electronic Fetal Monitoring.

A review of only those RCTs that involved high-risk pregnancies shows nonsignificant increases in the EFM group in the rates of perinatal death and of infants with Apgar scores of < 7 and nonsignificant decreases in the rates of neonatal seizures and admission to neonatal intensive care units. A statistically significant doubling of the rate of cesarean delivery was evident in the EFM group, however, and total operative deliveries were increased in high-risk pregnancies.

In pooling data from the trials in which fetal scalp-blood sampling was used to complement electronic monitoring, I found lower but statistically significant increases in the rates of cesarean delivery and total operative deliveries in the EFM group. The rate of neonatal seizure, on the other hand, decreased twofold when fetal scalp-blood sampling was used to complement electronic monitoring, although without the large Dublin trial this decrease is not statistically significant.

These findings leave pregnant women, as well as physicians and midwives, in a dilemma. Is preventing potentially serious but uncommon events (e.g., neonatal seizures) worth the much higher risk of operative delivery? The long-term clinical implications of seizures remain unclear. Published reports indicate that neonatal seizures are a serious prognostic finding; yet follow-up of the 39 Dublin infants with seizures who survived the neonatal period shows no difference in the outcomes for the auscultation and EFM groups at 1 year of age.³⁰ This finding suggests that more sensitive diagnostic criteria may have been used in the Dublin trial or that seizures with adverse prognostic implications are not affected by early intervention during labor and delivery. Investigators in the Dublin trial analyzed their data and found that the benefits of EFM were restricted to an association with protracted labor and with women given oxytocin (to induce or speed up labor). Although appropriately cautious about overanalyzing their data, they state that selective use of EFM may be preferable to universal use. The use of selective monitoring was tested in the Dallas and Seattle trials, but investigators found no measurable benefit to this approach, at least with their particular sets of selection criteria.^{31, 32}

Establishing policy in clinical practice requires not only determining whether a screening procedure is effective but also determining whether the human and monetary costs are acceptable. Moreover, the effectiveness, safety, and acceptability of subsequent interventions must be considered. For example, the evidence reviewed here demonstrates a significant increase in the rate of operative deliveries associated with EFM, with or without fetal scalp-blood sampling. If used in essentially all pregnancies, EFM will have large, direct financial costs, and these costs will increase dramatically if EFM is associated with an intervention (for example, cesarean delivery) that may often be unnecessary. When a decision is made about the routine use of EFM, its value should be assessed in view of the potential benefit of alternative obstetrical practices designed to decrease perinatal morbidity and mortality. Knowing these alternatives helps pregnant women, physicians, and midwives understand the impact of their choices on both maternal and infant well-being.

Diffusion of Technology

Diffusion refers to the spread of an innovation over time in a social system. The determinants of diffusion are complex. Fineberg ³³ has identified 10 influences on diffusion: (1) prevailing theory, (2) benefits of the innovation, (3) features of the clinical situation, (4) presence of an advocate, (5) characteristics of the adoption, (6) practice setting, (7) channels of communication, (8) the decision-making process, (9) evaluation, and (10) environmental constraints.

The perceived importance of asphyxia as a cause of neonatal mortality and morbidity in the 1970s facilitated the diffusion of EFM. Further, the benefits to the physician were great: EFM was relatively easy to learn, imposed little change on practice style, and replaced a practice—intermittent auscultation—that was imperfect and dependent on the skills of the nursing staff. EFM addressed a problem of great concern at the time—perinatal mortality—and appeared during a period of wide acceptance of new technology. It had strong advocates who were well represented in the obstetrical community, both in the United States and internationally.

The potential users of EFM—clinicians—were led by their colleagues in academic centers, who were at the forefront of EFM use and who communicated their preference in medical journals and at professional meetings. Newly certified obstetricians were uncomfortable with intermittent auscultation because as house officers they had had little experience in using it. Decision making in medicine moves most quickly when practice decisions are made exclusively by the individual practitioner, which probably facilitated the rapid spread of EFM use in the United States. All of these influences in the case of EFM tended to facilitate its rapid diffusion into clinical practice.

The remaining two sources of influence—environmental constraints through regulatory agencies and medical care insurers and evaluation through technology assessment—are influenced more by policymakers than physicians. In the sections that follow I address specific policies for EFM, as well as the role of technology assessment in general.

Impact of Technology Assessment

Although the impact of technology assessment on the use of EFM has not been quantified, it is clear that the initial diffusion of EFM was not affected by technology assessment. By the time the first assessment was published in 1979, nearly half of all deliveries in the United States were monitored electronically. In many academic centers the policy at that time was one of universal monitoring. In most hospitals, at least high-risk pregnancies were monitored electronically. The key data in assessing the effectiveness of this technology have come from RCTs. As noted previously, by 1987 nine RCTs had been conducted, but the impact of their results on clinical practice has been limited.

I cannot therefore estimate the impact on EFM of either the increasingly available data on efficacy and safety or formal technology assessment. There is little information, in fact, on the impact of the technology assessment process. The rapid implementation and discontinuation of the Swine Influenza Immunization Program illustrate the dramatic impact that social and political forces can have on the use of technologies.⁴² Historically, however, technologies tend to diffuse and disappear slowly, unless there are dramatic circumstances that force an action. The Methods and Evaluation Panels of the Council on Health Care Technology of the Institute of Medicine intend to examine evidence measuring the effectiveness of the technology assessment process.

Influence of Medical Malpractice on the Use of Electronic Fetal Monitoring

There is no doubt that many obstetricians have been encouraged to use EFM because of a fear of liability for not using the "customary procedure." The precise impact of malpractice concerns on the diffusion of EFM, however, has not been measured. More important, this fear of liability may not be well grounded. Careful reading of the relevant legal literature indicates that failure to use EFM should not result in liability, whereas using EFM in a routine labor and delivery may result in malpractice allegations.^{43, 44}

The critical legal assumptions regarding liability for not using a procedure are that (a) the procedure provides accurate and reliable information; (b) the information is of value for diagnosis; (c) the effective intervention is feasible following diagnosis; and (d) the procedure is better than other alternatives, in terms not only of effectiveness but also of safety. As noted, however, there is no consensus on these assumptions for EFM. As a result, physicians may not be liable for failing to use EFM, and those who do use EFM may be liable for failing to "keep abreast" or to "use best judgment," or even for ''negligence."

In medical malpractice the plaintiff must prove that an injury is the result of the physician's failure to act with "reasonable care." To establish causation, the use (or nonuse) of a technology or procedure must be shown to be the proximate cause of the injury; that is, nonuse (or use) is likely to reduce the risk of injury. If the allegation is that EFM should have been used, it must be shown that the use of EFM would have reduced the risk of injury. If, on the other hand, the allegation is that auscultation, not EFM, should have been used, it must be shown that EFM use significantly contributed to injury. Under the ordinary standards of negligence, as opposed to negligence standards applied to the medical profession, liability is found more often for using EFM, a monitoring procedure that entails greater risk than auscultation. Even under medical negligence standards, the physician needs only to show that the monitoring technique used (whether EFM or auscultation) adhered to customary practice.

The standards for "reasonable care" in medical liability are often associated with "customary practice." Does the physician possess and employ knowledge and skills in a reasonable manner, comparable to his or her peers? The legal interpretation of liability, however, is not limited to customary practice. In "ordinary negligence," failure to use a safe procedure could entail an inexcusable risk, beyond general standards of reasonableness. A whole industry (or specialty) may be found negligent for failing to adopt a new or safer technology or for prematurely using a new technology. Examples of premature diffusion of technologies in perinatal medicine include the use of high concentrations of oxygen in premature infants (leading to retrolental fibroplasia);⁴⁵ prescribing diethylstilbestrol (DES) for pregnant women to prevent miscarriages (leading to vaginal cancer in the children);⁴⁶ and prescribing a Dalkon shield, an intrauterine birth control device that was associated with septic spontaneous abortions and pelvic inflammatory disease.⁴⁷ Legal decisions in these cases were often based on the physician's duty to keep abreast of scientific knowledge and use the ''best judgment" based on that knowledge. Failure to do what a physician knows should be done can result in liability for an unfavorable outcome or injury.

In the case of EFM, because there is no consensus on the efficacy of the procedure and because there are risks (such as cesarean delivery) and costs associated with the practice, there appears to be no universally accepted standard of customary practice. Hence, the use of EFM in a particular case may not be justified.

The law recognizes a wide scope of discretion in the medical profession, and in a situation in which a "reputable minority" favors a particular practice, such as intermittent auscultation, no liability may be found when a physician fails to use a procedure favored by the majority (that is, customary practice). Use of EFM, on the other hand, may not protect the physician against liability in a suit brought because of complications arising from cesarean delivery when the use of an acceptable alternative (intermittent auscultation) was not likely to have led to a cesarean delivery.

The determination of liability will vary with the jurisdiction. "The legal standards employed to determine liability by courts in every jurisdiction do not provide a simple prescription for avoiding malpractice liability. Instead, physicians are required to use sound and reasonable judgment under the circumstances."⁴⁸

What I have described here is how the law is intended to work. In practice, however, decisions are often based on perceptions, both of plaintiffs and defendants. When a child is left with a serious disability, for example, it is difficult not to try to compensate the family and, as a result, find fault with the physician, even if the scientific evidence does not show negligence or incompetence. It is not surprising, therefore, that a 1987 survey conducted by the American College of Obstetricians and Gynecologists of about 2,000 of its members found that most claims are settled out of court.⁴⁹ Although EFM was not the focus of the survey, there were a few related findings. First, brain damage of the infant was significantly more likely to be the primary allegation in an obstetrical claim (31 percent) than any other category of primary allegation. In obstetrical claims the use of EFM was present to a significantly higher degree (46 percent) than any other characteristic. No specific details were provided on EFM-related litigation.

Given these circumstances, what can the obstetrician do to protect against malpractice claims? From the legal perspective, the best protection is informed consent.⁵⁰ A person has the legal right to make informed choices, and a well-informed patient is less likely to sue a physician. Moreover, because no technology or procedure can guarantee a perfect outcome, informing a patient will help to avoid unrealistic expectations. There is, of course, the difficulty of providing the patient with complete and unbiased information. A 1975 study of obstetrical and gynecological malpractice verdicts found that the "medical consumer frequently looks back upon this [the informed consent process] as 'selling the procedure' rather than giving information and getting consent."⁵¹ It is incumbent on the physician, therefore, to conscientiously provide a thorough, clear presentation of a procedure, including both the benefits and the risks. This practice will not only foster better physician-patient interaction but will serve the physician well in the event that mother and child suffer an injury or other adverse outcome.

Electronic fetal monitoring was introduced at a time when the obstetrician's primary concern was shifting from the mother to the fetus and newborn child. The 1970s were also a time of increasing use of technology in obstetrics and other areas of medicine. The obstetrician, motivated by a desire to protect the unborn child, was offered a variety of new tools—all promising not only to deliver more information but also to improve the outcomes of labor and delivery. As a consequence, EFM diffused rapidly, and its use has become standard medical practice throughout the United States. Unlike most technologies, EFM underwent a formal technology assessment—but only after it had become a standard practice. The impact of technology assessment and medical malpractice on the diffusion of EFM is not clear, but it was probably minor compared with the impact of the other factors governing diffusion. Practicing defensive medicine because of the fear of litigation may have a greater impact on the continued use of EFM in clinical practice.

It is interesting to note that the obstetrical community continues to debate the appropriate use of EFM. In April 1988 a committee of the American College of Obstetricians and Gynecologists recommended that the college endorse the position that EFM remains a useful tool but that even in high-risk pregnancies monitoring by auscultation is acceptable clinical practice.⁵²

References

1.

Banta, H. D., and S. B. Thacker. 1979. Costs and Benefits of Electronic Fetal Monitoring: A Review of the Literature. DHEW Pub. no. (PHS) 79-3245. Hyattsville, Md.: National Center for Health Services Research.

2.

Hon, E. H., and O. W. Hess. 1957. Instrumentation of fetal electrocardiography. Science 125:553-554. [PubMed: 13409008]

3.

Hon, E. H. 1959. The fetal heart rate patterns preceding death in uterus. Am. J. Obstet. Gynecol. 78:47-56. [PubMed: 13661244]

4.

Hon, E. H. 1960. Apparatus for continuous monitoring of the fetal heart rate. Yale J. Biol. Med. 32:397-399. [PMC free article: PMC2604143] [PubMed: 21434274]

5.

Hehre, F. W. 1974. Biophysical monitoring by fetal electrocardiography. Clin. Anesth. 10:81-101. [PubMed: 4590566]

6.

Hon, E. H., R. H. Paul, and R. W. Hon. 1972. Electronic evaluation of FHR. XI. Description of a spiral electrode. Obstet. Gynecol. 40:362-365. [PubMed: 5054966]

7.

Hehre. 1974; see note 5.

8.

Saling, E. 1961. Neue Untersuch Ungsmoglichkeiten des Kindes Unter Geburt (Einfuhrung and Grundlagen). Zent. Gynäkol. 83:1906-1908.

9.

Dilts, P. V. 1976. Current practices in antepartum and intrapartum fetal monitoring. Am. J. Obstet. Gynecol. 126:491-494. [PubMed: 984111]

10.

Paul, R. H., and E. H. Hon. 1970. A clinical fetal monitor. Obstet. Gynecol. 35:161-169. [PubMed: 5467078]

11.

Heldford, A. J., C. N. Walker, and M. E. Wade. 1976. Do we need fetal monitoring in a community hospital? Trans. Pac. Coast Obstet. Gynecol. Soc. 43:25-30. [PubMed: 982501]

12.

Lee, W. K., and M. S. Baggish. 1976. The effect of unselected intrapartum fetal monitoring. Obstet. Gynecol. 47:516-520. [PubMed: 1264397]

13.

Zdeb, M. S., and V. M. Logrillo. 1979. Prenatal monitoring in upstate New York. Am. J. Public Health 69:499-501. [PMC free article: PMC1619124] [PubMed: 434282]

14.

Placek, P. J., K. G. Keppel, S. M. Taffel, and T. L. Liss. 1984. Electronic fetal monitoring in relation to cesarean section delivery for live births and still births in the U.S. Public Health Rep. 99:173-183. [PMC free article: PMC1424552] [PubMed: 6424166]

15.

Banta, H. D., and S. B. Thacker. 1979. Assessing the costs and benefits of electronic fetal monitoring. Obstet. Gynecol. Survey 34:627-642. [PubMed: 113713]

16.

Thacker, S. B. 1987. The efficacy of intrapartum electronic fetal monitoring. Am. J. Obstet. Gynecol. 156:24-30. [PubMed: 3541620]

17.

Shy, K. K., E. B. Larson, and D. A. Luthy. 1987. Evaluating a new technology: The effectiveness of electronic fetal heart rate monitoring. Ann. Rev. Public Health 8:165-190. [PubMed: 3555520]

18.

Banta, H. D., C. J. Behney, and J. S. Willems. 1981. Toward Rational Technology in Medicine. New York: Springer. [PubMed: 10309422]

19.

Renou, P., A. Chang, I. Anderson, and C. Wood. 1976. Interpretation of the continuous fetal heart rate monitor. Obstet. Gynecol. 126:470-476. [PubMed: 10731]

20.

Haverkamp, A. D., M. Orleans, S. Langendoerfer, J. McFee, J. Murphy, and H. E. Thompson. 1979. A controlled trial of the differential effects of intrapartum fetal monitoring. Am. J. Obstet. Gynecol. 134:399-412. [PubMed: 453276]

21.

MacDonald, D., A. Grant, M. Sheridan-Pereira, P. Boylan, and I. Chalmers. 1985. The Dublin randomized controlled trial of intrapartum fetal heart rate monitoring. Am. J. Obstet. Gynecol. 152:524-539. [PubMed: 3893132]

22.

Renou et al. 1976; see note 19.

23.

Wood, C., P. Renou, J. Oates, E. Farrel, N. Beischer, and I. Anderson. 1981. A controlled trial of fetal heart rate monitoring in a low-risk obstetric population. Am. J. Obstet. Gynecol. 141:527-534. [PubMed: 7294080]

24.

Haverkamp et al. 1979; see note 20.

25.

Neldman, S., M. Oster, P. K. Hansen, J. Nim, S. F. Smith, and J. Hertel. 1986. Intrapartum fetal heart rate monitoring in a combined low-and high-risk population: A controlled clinical trial. Eur. J. Obstet. Gynecol. Reprod. Biol. 12:1-11. [PubMed: 3536631]

26.

MacDonald et al. 1985; see note 21.

27.

Haverkamp, A. D., H. E. Thompson, J. G. McFee, and C. Cetrulo. 1976. The evaluation of continuous fetal heart rate monitoring in high-risk pregnancy. Am. J. Obstet. Gynecol. 125:310-317. [PubMed: 5895]

28.

Kelso, I. M., R. J. Parsons, G. F. Lawrence, S. S. Arora, D. K. Edmonds, and I. D. Cooke. 1978. An assessment of continuous fetal heart rate monitoring in labor. Am. J. Obstet. Gynecol. 131:526-532. [PubMed: 677195]

29.

Leveno, K. S., F. G. Cunningham, S. Nelson, M. Roark, M. L. Williams, D. Guzick, S. Dowling, C. R. Rosenfeld, and A. Buckley. 1986. A prospective comparison of selective and universal electronic fetal monitoring in 34,995 pregnancies. N. Eng. J. Med. 315:615-619. [PubMed: 3736600]

30.

MacDonald et al. 1985; see note 21.

31.

Leveno et al. 1986; see note 29.

32.

Luthy, D. A., K. K. Shy, G. Van Bell, E. B. Larson, J. P. Hughes, T. J. Benedetti, Z. A. Brown, S. Effer, J. F. King, and M. A. Stenchever. 1987. A randomized trial of electronic fetal monitoring in premature labor. Obstet. Gynecol. 69:687-695. [PubMed: 3554055]

33.

Fineberg, H. F. 1985. Effects of clinical evaluation on the diffusion of medical technology. Pp. 176-210 in Assessing Medical Technologies. Washington, D.C.: National Academy Press.

34.

Banta, H. D., and S. B. Thacker. 1979. Policies toward medical technology: The case of electronic fetal monitoring. Am. J. Public Health 69:931-935. [PMC free article: PMC1619228] [PubMed: 474850]

35.

Haverkamp et al. 1979; see note 20.

36.

Haverkamp et al. 1976; see note 27.

37.

National Institute of Child Health and Human Development. 1979. Part III: Predictors of Fetal Distress. I. Antenatal Diagnosis. NIH Pub. no. 79-1973:1-199. Washington, D.C.: Government Printing Office.

38.

Kosecoff, J., D. E. Kanouse, W. H. Rogers, L. McCloskey, C. M. Winslow, and R. H. Brook. 1987. Effects of the National Institutes of Health Consensus Development Conference on physician practice. JAMA 258:2708-2713. [PubMed: 3499522]

39.

Pub. L. No. 94-295.

40.

Anderson, C. G. 1975. Monitoring in labor, a patient cost survey. Contemp. Obstet. Gynecol. 6:102-104.

41.

Dans, P. E., J. P. Weiner, and S. E. Otter. 1985. Peer Review Organizations: Promises and potential pitfalls. N. Eng. J. Med. 313:1131-1137. [PubMed: 3930964]

42.

Neustadt, R. E., and H. V. Fineberg. 1978. The Swine Flu Affair: Decision-making on a Slippery Disease. Washington, D.C.: U.S. Department of Health, Education, and Welfare. [PubMed: 25032342]

43.

Gilfix, M. G. 1984. Electronic fetal monitoring: Physician liability and informed consent. Am. J. Law Med. 10:31-90. [PubMed: 6391159]

44.

Katz, B. F. 1979. Electronic fetal monitoring and the law. Birth Fam. J. 6:251-258.

45.

Silverman, W. 1980. Retrolental Fibroplasia: A Modern Parable. Orlando, Fla.: Grune and Stratton.

46.

Gunning, J. E. 1976. The DES story. Obstet. Gynecol. Survey 31:827-833. [PubMed: 980278]

47.

Layde, P. M. 1983. Pelvic inflammatory disease and the Dalkon shield. JAMA 250:796-797. [PubMed: 6876335]

48.

Gilfix. 1984; see note 43.

49.

American College of Obstetricians and Gynecologists. 1988. Professional Liability and Its Effects: Report of a 1987 Survey of ACOG's Membership. Washington, D.C.

50.

Gilfix. 1984; see note 43.

51.

Shearer, M., M. Raphael, and M. Cattani. 1976. A survey of California OB-GYN malpractice verdicts in 1974 with recommendations for expediting informed consent. Birth Fam. J. 3:59, 64.

52.

Cogen, J. 1988. ACOG considers new guidelines for monitoring and labor. Ob/Gyn News 23:1, 43.

Footnotes

*

Because the Dallas study was designed to compare the results of hospital policies that related to the restricted availability of EFM rather than the use of EFM compared with auscultation, I did not pool these data.²⁹

Use of trade names is for identification only and does not constitute endorsement by the Public Health Service or the U.S. Department of Health and Human Services.