Please note: In 2003, the CTF updated its Grades of Recommendations to include an "I Recommendation" for situations where insufficient evidence exists to allow a recommendation to be made. (Formerly, these situations were captured under a "C Recommendation".) This change is not retroactive, and all "C Recommendations" made prior to 2003 have not been reevaluated in light of the new "I" recommendation grade. For a discussion of these recommendation grades, please link to the 2003 article in the Canadian Medical Association Journal here.
These recommendations were finalized by the Task Force in June 1993
Contents
This review of the evidence for the use of home uterine activity monitoring (HUAM) in the prevention of preterm birth is based on the U.S. Preventive Services Task Force review of the subject.<1> The review identified 6 randomized trials of HUAM screening and subsequent therapeutic interventions that reported perinatal outcomes. There are no controlled trials of home uterine activity monitoring (HUAM) in pregnancies with no risk factors for preterm labour. All of the trials included populations of women at high risk for preterm labour.
The Canadian Task Force concurs with the U.S. Preventive Services Task Force that the lack of evidence of benefit of HUAM in normal pregnancies suggests that this maneuver not be included in the management of normal pregnancies (D Recommendation).
The results of 6 randomized controlled trials of HUAM screening in women with high risk of preterm labour have been published. Four of these indicated some positive effects of HUAM screening. However, the design of two studies made it difficult to separate the effects of HUAM from other components of the experimental intervention. The two other studies which indicated positive effects of HUAM screening used controversial techniques in the statistical analysis of the trial results. Although HUAM screening of high risk pregnancies may be a promising technique for the prevention of preterm birth, currently there is insufficient evidence of clinical effectiveness to recommend for or against its routine use as a screening maneuver. (C Recommendation)
Burden
of Suffering Maneuver Effectiveness
of Prevention and TreatmentThis review of the evidence on the impact of HUAM screening excludes studies that solely examine the value of HUAM in detecting preterm labour and is limited to studies that examine the impact of programs that combine detection of preterm labour using HUAM and the treatment of preterm labour with tocolytic agents.
The design of most trials of HUAM is similar. High-risk patients are selected on the basis of a history of prior preterm deliveries, multiple gestations, uterine anomalies, or other risk factors. Patients receive initial training and are instructed to monitor their contractions at home two to four times each day and to transmit stored data daily via telephone to receiving tocographs. Study personnel are available by telephone, sometimes on a 24-hour basis, to answer patient inquiries. If the patient detects more than four contractions per hour or other specified symptoms, she monitors contractions for an additional hour while lying on her side and taking fluids. If the symptoms persist, the patient is evaluated in hospital. If preterm labour does not resolve with rest and hydration, tocolytic therapy is begun. Long-term tocolysis, if instituted, is generally discontinued at 35-37 weeks. Patients are usually examined every one to two weeks. Monitoring is discontinued at 36-37 weeks. Preterm labour is defined by persistent uterine contractions and cervical change, and preterm delivery is usually defined as a delivery before 36 completed gestational weeks.
Morrison et al (1987)<4> assigned 69 out of 75 eligible high-risk patients to a study group that received HUAM, education, and self-palpation instructions or a control group that received education and self-palpation instructions only. Although monitored women were contacted daily, subjects in the control group were contacted by telephone twice each week.
There was no difference between groups in the incidence of preterm labour or in the gestational age at which preterm labour was diagnosed. Among women who developed preterm labour, there were statistically significant differences between groups in the incidence of cervical dilation of 3 cm or more, cervical effacement of greater than 50%, and rupture of membranes. A greater proportion of monitored cases were eligible for long-term tocolysis. The time gained in utero was an average of four weeks longer in the monitored group, and the proportion that reached a gestational age of 36 weeks was also significantly greater. Clinical outcomes in newborns were reported in a separate paper.<5> The authors reported a significantly lower overall number of complications in the monitored group than in controls.
Study and control groups differed in the intensity of nursing follow-up; women in the study group received daily telephone calls from study nurses, whereas women in the control group were contacted twice each week.
Iams et al<6> randomly assigned 157 high-risk women at 20-34 weeks gestation to a study group that received HUAM, education and self-palpation instructions only. In contrast to the previous study, women in the control group were contacted at least five times each week, daily on weekdays and as needed on weekends. Fifteen women dropped out of the study after randomization. There were no statistically significant differences between groups in the incidence of preterm labour, preterm deliveries, gestational age at delivery, or mean birth weight. Clinical outcomes in the newborn were not measured.
In 1988, Iams et al (1988)<7> published a subsequent report based on additional 152 subjects enrolled in the second year of the study, providing a total sample of 309 patients over two years. Twenty-eight subjects in the second year dropped out after randomization. Over the two-year period, the authors again found no significant differences between groups in preterm labour and delivery rates, mean birth weight, or gestational age at delivery.
The statistical power of this study, even with the larger sample, has been questioned by some critics but this has since been refuted by the authors.<8> Other criticisms include the lack of a standardized definition for preterm labour, the failure of some physicians in the study to continue tocolysis beyond 35 weeks, inadequate randomization, and the administration of prophylactic tocolytics by some physicians.<9>
In a multicentre study, Hill et al (1990)<10> randomly assigned 299 high-risk women at 20 to 34 weeks gestation to a study group that received daily HUAM, education, and daily nursing contacts or to a control group that received education and self-palpation instructions. Women in the control group were advised to contact their physicians for certain symptoms and did not receive frequent nursing contacts. Fifty-four women (18%) were excluded from the study following randomization. There was no significant difference between the groups in the incidence of preterm labour. The monitored group had significantly fewer cases of cervical dilation beyond 2 cm and delivery within 48 hours of the first labour episode (18% vs. 37%). Term delivery was achieved by more women in the monitored group than in the control group. Clinical outcomes in the newborn were not measured.
Women in the study group were contacted daily, whereas women in the control group were not contacted regularly. High attrition and exclusion rates may also have influenced the results.
Dyson et al (1991)<11> provided education, self-palpation instructions, and HUAM to 251 high-risk women in California and then randomized them to two groups. Both groups transmitted monitor data, but data transmitted by women in the control group were not analyzed or used in patient management. Patients were unaware of their group assignments. Results from both groups were compared to data from a non-randomly selected "standard care" group, consisting of 143 women with similar risk factors who were cared for in the 30 month period before the study. All patients were contacted five days per week. Outcomes were reported separately for singleton and multiple gestations.
For singleton gestations, the incidence of preterm births was lower in the control group than in the standard care group, but the rate of preterm births in the monitored group did not differ significantly from outcomes in the control group. Among women who developed preterm labour before 34 weeks, outcomes were poorer in the standard care group than in the monitored and control groups.
In twin gestations, the incidence of preterm birth was significantly lower in the monitored and control groups than in the standard care group. There were also significantly lower rates of neonatal death, intensive care nursery admission, and respiratory distress syndrome in the control group than in the standard care group. All neonatal outcomes were significantly better in the monitored group than in the standard care group. The monitored group had significantly lower rates than controls in the incidence of birth weight less than 1,500 g, intensive care nursery admission, and length of hospital stay.
This study benefitted from a large sample size and an innovative design that attempted to separate the effects of HUAM from those of nursing contact. Unfortunately, overall results were not reported to determine whether monitored women as a group, including those with singleton and twin gestations experienced better outcomes than women in the control group. Significant results for women with twin gestations were reported for three out of eight clinical outcomes, but the calculations of statistical significance were based on different denominators. Although calculations for the singleton group used the number of women as the denominator, calculations for twin gestations used the number of newborns as the denominator. This doubling of the denominator increased the apparent degrees of freedom, thereby achieving statistical significance. The use of newborns as the denominator for twin gestations is of concern because newborns were not the unit of analysis in the study and because twin gestations are not independent variables.
In a multicentre study, Mou et al (1991)<12> randomly assigned 377 of 509 eligible high-risk women at 23-30 weeks gestation to a study group that received HUAM, education, and self-palpation instructions or to a control group that received education and self-palpation instructions. The 132 women who were not enrolled (26% of the eligible sample) failed to participate because they refused, because consent was not requested, or because they lacked a telephone. Women in the control group were not contacted by study personnel, and personnel who contacted women in the study group were instructed to report the number of contractions but not to ask questions or provide medical advice.
There was no significant difference between groups in the incidence of preterm labour or in the gestational age at the diagnosis of preterm labour. Although clinical outcomes were examined, cervical dilation was the primary end point of the study. When the number of women who developed preterm labour was used as the denominator, mean cervical dilation was significantly similar in the monitored group (1.4 cm) than in the control group (2.5 cm). The proportion with cervical dilation less than 2 cm, time gained in utero after diagnosis of preterm labour, and mean gestational age at delivery were also significantly greater in monitored women than in controls. Among a subset of singleton gestations (see below), newborns from monitored pregnancies had significantly greater birth weight (2,934 g vs. 2,329 g) and required fewer days of neonatal intensive care (50 vs. 324 days), oxygen therapy (0 vs. 68 days), and mechanical ventilation (0 vs. 54 days) than newborns from unmonitored pregnancies.
The randomization did not include 132 (26%) of the 509 eligible women, and about 11% of the randomized sample were subsequently excluded. Of the remaining 334 women, 252 women who did not develop preterm labour (67% of the randomized sample) were not included in the overall calculations. The major findings of the study were drawn from the remaining 82 women who developed preterm labour. In this group, monitoring was associated with a significantly smaller cervical dilation at the onset of labour, but it is unclear whether this difference in dilation resulted in improved clinical outcome for the newborns. Although significant differences between the groups were reported for gestational age at delivery, birth weight, neonatal intensive care, and respiratory therapy, the calculations were based on a subset of the 82 women, excluding 19 (23%) cases that involved twin gestations and medically indicated preterm deliveries.
In a French study, Blondel et al (1992)<13> randomly assigned 94 high-risk women to a study group that received HUAM and daily midwife contact or to a control group that received home visits by a community health nurse once or twice each week. The monitoring device standards, and organization of home care were similar to those used in the U.S. studies. The patients were at 24-34 weeks gestation and were recruited from four public maternity units.
The investigators found no difference between groups in the rate of hospital admission for threatened preterm labour. There were no statistically significant differences between groups in the incidence of preterm births before 37 weeks or of birth weight less than 2,500 g. No other clinical outcomes were measured.
Although this evidence suggests at least some perinatal benefit from HUAM screening there are important conceptual or methodological problems with each of the positive trials.<4,5,10-12> In two of the earlier studies<4,5,10> the experimental and control groups differed not only in the use of HUAM, but also in the level and intensity of contact with nursing staff. This makes it difficult at a conceptual level to determine if the difference in observed outcomes is the result of HUAM screening or nursing contact. The two latter trials<11,12> that have positive results do not have a problem with defining HUAM as the intervention, but involve the use of controversial statistical analysis techniques.
Recommendations
of OthersAt this time the Society of Obstetricians and Gynaecologists of Canada has not made an official statement on guidelines for the use of HUAM.
Conclusions
and RecommendationsHUAM screening could cost between $600 and $7,000 per monitored pregnancy. Given the high cost of this maneuver and the lack of any experimental evidence of its benefits in normal pregnancies, its routine use in normal pregnancies cannot be justified at this time (D Recommendation). The problems with existing trials in high-risk pregnancies means that there is inconclusive evidence regarding the inclusion or exclusion of HUAM in the high-risk population (C Recommendation).
Unanswered
Questions (Research Agenda) EvidenceThis review was initiated in March 1993 and recommendations finalized by the Task Force in June 1993. For a complete reference list please see reference 1.
Link to Structured Abstract of this review
Link to Summary Table of this review
Link to Selected References list of this review
Reprinted in modified format by the Canadian
Task Force on Preventive Health Care
with permission.
For any technical issues please contact: webmaster@ctfphc.org
Original Copyright
© 1994 Minister of Supply and Services Canada.
Last modified March 27, 1998.
