Rapid Responses to:
|
|
Rapid Responses: Rapid Responses published:
-
![[Read Rapid Response]](/icons/misc/sectionDOWN.gif)
What do women think?
- Hannah Hulme Hunter (15 June 2001)
-
Importance of controlling for potential confounding
- Deirdre J Murphy (16 June 2001)
-
Estimating the effects of health care interventions
- Iain Chalmers (16 June 2001)
-
Is proper statistics a religion?
- Izhar Ben-Shlomo (16 June 2001)
-
Statistical or clinical significance?
- Simon Grant (18 June 2001)
-
Failure to achieve expected sample size should not prevent publication of randomised trials
- Francesco Torella (20 June 2001)
-
Original power calculations may be flawed
- Paul Little (25 June 2001)
-
How should we interpret randomized controlled trials (RCT), which compare unreproducible methods?
- J Bernades, A Costa-Pereira (27 June 2001)
-
Reducing obstetric morbidity in low risk women
- Lesley Dolan (28 June 2001)
-
Doppler auscultation standard not evidence based
- Jane Grant (29 June 2001)
-
Unblinded studies and source of bias
- Alison Macarthur (3 July 2001)
-
Changing risks: what about the neonate?
- W M Clow (11 July 2001)
-
Questions about hypoxia and acidosis
- Shane Higgins (17 July 2001)
-
Electronic Fetal Monitoring
- Jane Thomas (19 July 2001)
-
Abnormal neonatal neurobehaviour may be a more sensitive indicator of intrapartum asphyxia than cord
- Margaret Sheridan-Pereira, Adrian Grant (23 July 2001)
-
Recruitment for clinical trial:the key to success
- Paola Albertazzi (13 August 2001)
|
|
|||
|
Hannah Hulme Hunter, midwife Maternity Unit, Stoke Mandeville Hospital, Bucks.
Send response to journal:
|
Were women asked how they felt about admission CTGs? Does this intervention reassure or terrify? Is it a gross intrusion into the flow of labour, a minor irritation, or one of the reasons why they choose to give birth in hospital? Is this study yet another missed opportunity to hear the voices of the women for whom we care? |
|||
|
|
|||
|
Deirdre J Murphy, Consultant Senior Lecturer in Maternal Medicine St Michael's Hospital, Bristol
Send response to journal:
|
Mires et al report an increased rate of operative delivery in women who had admission cardiotocography and suggest that this is perhaps the most important finding of their study. (1) This statement is likely to generate a lot of heated debate particularly among groups committed to reducing intervention rates in obstetric and midwifery practice. It is important that this statement has a firm basis and is not interpreted out of context. The authors report an association between admission cardiotocography and operative delivery of OR 1.36, (95% CI 1.12,1.65). They also note that admission cardiotocography was associated with an increased rate of continuous fetal heart rate monitoring in labour OR 1.49 (1.26,1.76). It has long been understood that continuous electronic fetal heart rate monitoring is associated with an increase in caesarean and operative vaginal delivery RR 1.41, (95% CI 1.23,1.61) and RR 1.20 (1.11,1.30) respectively (2). The authors suggest that the increase in continuous monitoring is likely to contribute to the reported association. I suspect that the increase in continuous monitoring may fully explain this apparent association. A true estimate of the association between admission cardiotocography and operative delivery could be represented by an odds ratio adjusted for the potential confounding effect of continuous monitoring. The important contribution of continuous monitoring to this association is represented by the stronger association between continuous monitoring and operative delivery than that of admission cardiotocography and by the consistent association between continuous monitoring and increased operative delivery reported in previous trials. (2) The authors have all the necessary data to perform the proposed analysis which would be helpful in clarifying an issue that has important clinical implications. The association between admission cardiotocography and increased continuous fetal heart rate monitoring will have important consequences but this pathway needs to be clearly reported and interpreted. (1) Mires G, Williams F, Howie P. Randomised controlled trial of cardiotocography versus Doppler auscultation of fetal hear at admission in labour in low risk obsteric population. BMJ 2001;322:1457-62. (2) Thacker SB, Stroup D, Chang M. Continuous electronic heart rate monitoring for fetal assessment during labor (Cochrane Review). In: The Cochrane Library, Issue 2. Oxford:Update Software. |
|||
|
|
|||
|
Iain Chalmers, Director UK Cochrane Centre
Send response to journal:
|
Sandy Goldbeck-Wood’s invitation (1) to respond to questions raised by the first randomised comparison of cardiotocography versus auscultation of the fetal heart at admission in labour prompts me to reiterate the following points. The term “negative trial” should be outlawed (2), as should editorial acquiescence in statements implying that it is possible to prove negatives. For example, the BMJ report of the admission cardiotocography trial states that this intervention “does not benefit neonatal outcome in low risk women”, yet the estimate of its effect on the incidence of hypoxic ischaemic encephalopathy is compatible with the reduction in neonatal seizures found in other RCTs of fetal monitoring during labour (3). Although it is worth making pre-trial estimates of the sample sizes needed to rule in or rule out, with specified statistical confidence, effects of a magnitude likely to be relevant to patients, how many participants you needed in a trial depends on what you found, not on what you thought you would find (4). Statements that it is unethical to embark on controlled trials unless an arbitrarily defined level of statistical power can be assured make no sense if the alternative is acquiescence in ignorance of the effects of healthcare interventions. Unbiased estimates should be generated through randomised trials because they leave us less ignorant than we were before (5). The importance of the admission cardiotocography study is that it provides the first unbiased estimates of the ranges within which the effects of this healthcare intervention are likely to lie. All unbiased comparative studies should be published, so that the totality of the relevant evidence can be evaluated in the systematic reviews that are needed by those making choices and decisions in health care, and about further research. It is now possible to register and publish successive versions of trial protocols and all subsequent reports of registered studies(6). Reluctance to exploit this potential does not reflect scientific or technical problems, but rather academic and commercial interests. References 1. Goldbeck-Wood S. Commentary: changes between protocol and manuscript should be declared at submission. BMJ 2001;322:1460-1. 2. Chalmers I. Proposal to outlaw the term 'negative trial'. BMJ 1985;290:1002. 3. Thacker SB, Stroup D, Chang M. Continuous electronic heart rate monitoring for fetal assessment during labor (Cochrane Review). In: The Cochrane Library, Issue 2, 2001. Oxford: Update Software. 4. Detsky AS, Sackett DL. When was a "negative" clinical trial big enough? How many patients you needed depends on what you found. Arch Intern Med 1985;145:709-12. 5. Lilford RJ, Thornton JG, Braunholtz D. Clinical trials and rare diseases: a way out of a conundrum. BMJ 1995;311:1621-5. 6. Chalmers I, Altman DG. How can medical journals help prevent poor medical research? Some opportunities presented by electronic publishing. Lancet 1999;353:490-493. |
|||
|
|
|||
|
Izhar Ben-Shlomo, Director of Research Dept. Ob/Gyn, HaEmek Med. Ctr., Afula, Israel
Send response to journal:
|
Atheists believe that man has invented god. Believers object. Are the guidlines to proper statistical methodology in clinical trials, as presented religiously by Murray, a preexisting god, or should they be regarded as a man-made one? |
|||
|
|
|||
|
Simon Grant, Consultant O&G RCHT, Treliske, Truro
Send response to journal:
|
Dear Sir Mires et al conclude that admission cardiotocography (CTG) does not benefit neonatal outcome in low risk women, but that it does increase the rate of operative delivery(1). No information is given regarding the classification of the admission traces in their study. Other responses have addressed the likely cause of this increased operative delivery rate (2) and the suggestion that lack of effect can be proven (3). An earlier study, although not randomised, was a prospective blinded study of women designated as low risk on admission in labour and included a similar number of women to the CTG arm of the current study once defined as low risk at labour onset (1041 compared to 1186)(4). This study showed a 1% incidence of an "ominous" fetal heart rate (FHR) trace on admission testing and, of these ten infants, one died three hours after admission, without stethoscopic auscultation detecting any fetal compromise. This again highlights the point made by Chalmers (3). The conclusions drawn from the current study fail to recognise the difference between statistical significance and clinical significance in an individual case, particularly when the adverse outcome is rare. Intermittent auscultation will not detect the development of all FHR abnormalities and depends on the ability to provide adequate levels of midwifery care. The inability to take the CTG off after an admission test is as much a training issue as the inability to recognise abnormal the CTG identified by CESDI(5). References: 1) Mires G, Williams F, Howie. Randomised controlled trial of cardiotocography versus Doppler auscultation of fetal heart rate at admission in labour in low risk population. BMJ;322:1457-60, 2001. 2)Murphy DJ. Importance of controlling for potential confounding. eBMJ Rapid Responses, June 2001. 3)Chalmers I. Estimating the effects of health care interventions. eBMJ Rapid Responses, June 2001. 4)Ingemarsson I, et al. Admission test: a screening test for fetal distress in labor. Obstetrics & Gynecology. 68(6):800-6, 1986 Dec. 5)Confidential Enquiry into Stillbirths and Deaths in Infancy. 1st Annual Report. London: DoH, 1995. |
|||
|
|
|||
|
Francesco Torella, Research Fellow in Surgery South Manchester University Hospital
Send response to journal:
|
Dear Editor, I agree with the your decision to publish this study, as useful information is made available on a widely read journal. Pre-trial information may be scarce and too often power calculations are based on `wild` assumptions, which turn out to be unrealistic as the trial progresses. Editors of medical journals should encourage submission of trial protocols at the outset of any large study (maybe in a predefined format)to ensure clarity. Furthermore, when the initial target is not achieved, why not report both the initial power calculation and the actual power of the sample size at the end of recruitment? This should ensure transparency and should automatically require an eplaination for the failure to recruite enough patients. |
|||
|
|
|||
|
Paul Little, Senior Lecturer Primary Medical Care Group, Southampton University
Send response to journal:
|
Dear Editor, Dr Goldbeck-Wood encourages response to the issue of original power calculations. The BMJ is right to publish studies where the power calculation has changed, since original power calculations may well be based on arbitrary judgements or imprecise guestimates. Initial power calculations may be based on: 1) estimates from previous studies, which may differ significantly from the proposed study in terms of the characteristics of the population to be studied (in particular the prevalence of the main outcome for dichotomous outcomes, or the mean and standard deviation for continuous outcomes) 2) estimates based on a judgement of what is clinically significant: this is clearly not a precise science and what one clinician regards as significant another will not. This judgement may also change over time. Preferably these judgements should be made by multidisciplinary groups, but even then the validity of such judgement is questionable and arbitrary 3) surrogate endpoints. Better validated measures may be developed during the course of the trial. 4) incorrect or overly conservative assumptions Thus there may be good reasons to revise the sample size
calculation, particularly if more secure information becomes available
from the trial population or from other similar studies. To avoid the
accusation that such decisions are arbitrary and driven by the final
analysis, good practice should probably involve revised sample size
calculations being:
The BMJ is also right to encourage authors to share the original power calculations and the rationale and process of changing the calculations – and authors are more likely to do this if they feel that journal editors will not punish them for such openness. Yours Sincerely Paul Little
|
|||
|
|
|||
|
J Bernades, Professors Obstetrics&Gynecology and Biostatistics&Medical Informatics, Oporto Faculty of Medicine, Portugal, A Costa-Pereira
Send response to journal:
|
As stressed by Mires et al. 1 there is wide intra and inter-observer variation in cardiotocograph (CTG) interpretation. This means that there may be a wide variation in CTG sensitivity and false positive rate, which can influence its related efficacy and unnecessary obstetric interventions. For example, it can be demonstrated that for an observer variation with a proportion of agreement of 0.33 and 0.71 for abnormal and normal tracings, respectively, and an assumed sensitivity and false positive rate of 50% and 17%, respectively, the latter may vary from 0% to 100% and 0% to 33%, respectively (http://sisporto.med.up.pt/intraobserver.html). In the same way, it can also be demonstrated that for a proportion of agreement of 0.33 and 0.71 for intervention and no action, respectively, and an assumed efficacy and rate of unnecessary interventions of 50% and 17%, respectively, the latter may vary from 0% to 100% and 0% to 33%, respectively. This may, in different studies, lead to discrepancies on relative risks and odds ratios and may also lead to poor average results in multi-observer and meta-analysis studies (http://sisporto.med.up.pt/intraobserver.html).
How can we then interpret the results from studies on cardiotocography like the RCT conducted by Mires et al.? Were the differences due to the nature of the methods used or to how they were performed or to both problems? As we can not answer this question should cardiotocography be considered a method which has already been fully evaluated and be banned in low risk populations or should we instead recognize that cardiotocography has not yet been evaluated properly? In our opinion, the above problems can only be overcome if we focus the research on more objective CTG analysis 2-4 and then perform RCTs based on a reproducible method. 1- Mires G, Williams F, Howie P. Randomised controlled trial pf cardiotocography versus Doppler auscultation of fetal heart at admission in labour in low risk obstetric population. BMJ 2001;322:1457-62. 2- Ayres-de-Campos D, Bernardes J, Costa-Pereira A, Pereira-Leite L. Inconsistencies in expert’s classification of cardiotocograms and subsequent clinical decision. Br J Obstet Gynecol 1999;106:1307-10. 3- Bernardes J, Costa-Pereira A, van Geijn HP, Pereira-Leite L. A more objective fetal heart rate baseline estimation. Br J Obstet Gynaecol 1996;103:714-5. 4- Ayres-de-Campos D, Bernardes J, Garrido A, Marques-de-Sá, Pereira-Leite L. SisPorto 2.0 – a program for automated analysis of cardiotocograms. J Matern Fetal Med 2000;9:311-318. Bernardes and Costa-Pereira are involved in the development of a computer based system for CTG analysis (SisPorto). |
|||
|
|
|||
|
Lesley Dolan, Consultant Psychiatrist Murray Royal Hospital, Perth
Send response to journal:
|
EDITOR - Surely the most worrying finding of this study is that 31% of low risk women admitted to this teaching hospital in labour ended up with an operative delivery, irrespective of randomisation. This is a sad indictment of obstetric care for a low risk group. The real question is whether a teaching hospital is a safe place for low risk women to give birth. It is time to challenge the increasing centralisation of obstetric care in teaching hospitals which may well contribute to an unnecessary increase in morbidity for normal women. Lesley Dolan 1. Mires G, Williams F, Howie P. Randomised controlled trial of cardiotocography versus Doppler auscultation of fetal heart at admission in labour in low risk obstetric population.BMJ 2001; 322:1457-1462. |
|||
|
|
|||
|
Jane Grant, Research Fellow South Banlk University
Send response to journal:
|
In the description of their randomised controlled trial (RCT) protocol Mires et al write that low risk women were randomised to either 20 minutes of cardiotocography at admission in labour, or to ‘fetal heart rate auscultation with a hand held Doppler device during and immediately after at least one contraction’ (1) The authors reference the source of their CTG standard of 20 minutes (2), but give no rationale for their Doppler auscultation standard. The latter is surprising since the standard for intrapartum fetal heart auscultation in low risk women, as evaluated in RCTs, was published as long ago as 1993 (3) and again this year in the National Institute for Clinical Excellence guideline for the use of electronic fetal monitoring in labour (4). The recommendations are to auscultate the fetal heart after a contraction for a minimum of 60 seconds at a frequency determined by the woman’s stage in labour. There is, to date, no evidence to justify a standard of Doppler auscultation during and immediately after a contraction. The authors’ departure from the recommended standard is puzzling and potentially damaging if it undermines the implementation of appropriately evaluated recommendations (4).
Jane Grant, (1) Mires G, Williams F, Howie P. Randomised controlled trial of cardiotocography versus Doppler auscultation of fetal heart at admission in labour in low risk obstetric population. BMJ 2001;322:1457-62. (2) Ingermarsson I. Electronic fetal monitoring as a screening test. In: Spencer JAD, Ward RHT, ed. Intrapartum fetal surveillance. London: Royal College of Obstetricians and Gynaecologists, 1993:45-52. (3) Recommendations arising from the 26th RCOG Study Group: Intrapartum Fetal Surveillance. In: Spencer JAD, Ward RHT, ed. Intrapartum fetal surveillance. London: Royal College of Obstetricians and Gynaecologists, 1993:387-393. (4) National Institute for Clinical Excellence: The use of electronic fetal monitoring. Inherited Guideline C, 2001. |
|||
|
|
|||
|
Alison Macarthur, Associate Professor, Dept. of Anesthesia, University of Toronto Mount Sinai Hospital
Send response to journal:
|
Dear Drs. Mires, Williams and Howie; Thank you for the presentation of your study’s findings – as an obstetric anesthetist involved in a large, tertiary care obstetric practice this study’s findings may be relevant for only a small proportion of our patients. However, prior to my acceptance of the study’s applicability to low risk women presenting in spontaneous labor (who are the defined target population), I would like clarification of results not reported in the manuscript. First, would you please demonstrate that adequate randomisation occurred by detailing table 2’s information (reasons for exclusion after randomisation) for each of the two interventions (cardiotocograph or Doppler). We cannot ascertain whether, by chance alone, more women with need for continuous cardiotocographic monitoring were randomised to the initial cardiotocographic intervention. Certainly a selection bias was demonstrated when women receiving cardiotocographic intervention had a greater need for continuous monitoring during labor (21.5% vs. 3.6% in Doppler intervention group). If the randomisation process worked adequately for the two groups in distributing other associated factors equally then the intention-to-treat analysis is the superior result. I have ignored the sub-group analysis because of the concern that the intervention was unblinded and therefore could have affected the decision for continuous fetal heart monitoring because of study allocation. I believe the requirement for continuous monitoring during labor is an important factor associated with increased incidences of augmented labor, artificial rupture of membranes, and increased epidural use. However, the association must then be shown to be causal. Also, each of these intervention–outcomes risk associations had the lower 95% C.I of the relative risk include 1.00 (table 1) and therefore where not significantly increased in either intervention group. The other issue, as raised in the accompanying editorial, is that rare neonatal morbidity outcomes are difficult to evaluate in single center clinical trials. This study had insufficient power to demonstrate that fetal heart rate assessment on admission by Doppler was not associated with increased deleterious intrapartum sequelae to the neonate. (Goddard R. Electronic fetal monitoring. BMJ 2001; 322: 1437- 7.) Insufficient power does not allow us to dismiss the possibility that cardiotocographic initial assessment of fetal heart rate may better identify early fetal intolerance of the stress of labor. The harder leap of faith to make is that the method of assessment used for triage assessment of fetal heart rate is an associated factor (I would not suggest causation as it requires even further evidence by Hill’s criteria) for operative delivery. I would appreciate hearing from the authors the usual sequence of events that led to the decision of operative vaginal delivery (as the cesarean delivery rate was essentially equivalent between intervention groups). Was the delivering health professional in an institution where the presence of an epidural for labor analgesia has been shown to increase the use of vacuum or forcep-assisted deliveries, inorder to teach training physicians?(Bofill JA et al. Nulliparous active labor, epidural analgesia and cesarean delivery for dystocia. Am J Obst Gyne 1997; 177:1465-70) Was there a limitation in the delivering professional’s time that may have led to an expedited operative vaginal delivery. Was there any attempt at blinding the individual who made the decision for operative delivery to the patient’s designated assignment in the study? All of these questions indicate potential sources of selection bias that could influence and / or exaggerate the true effect of the intervention on the outcome. I would like more information on those women whom the umbilical cord gas evaluation was not done or not adequate for the study. Obviously bias may have affected the primary outcome results depending upon the reasons for why gases where not obtained and the patient population that samples were missing from. The reasons for the one-third missing outcome information may indicate the direction of their effect on the relative risk (1003 / 3751 missing primary outcome information in intention-to- treat analysis). Lastly, I believe the “incidence of epidural analgesia for labor” as a negative secondary outcome to be out-dated. The issue of association between labor epidural analgesia and cesarean delivery has been well examined and discarded by recent excellent studies. (Halpern S et al. JAMA 1998; 280: 2105-10, Segal et al Am J Obst Gyne 2000; 183: 974-8, Loughnan BA et al, Br J Anaes 2000; 84: 115-9) The association between epidural analgesia and operative vaginal delivery rates appears more consistent in studies. However, whether this is a causal association due to change in pelvic floor function, or simply the effect of a comfortable parturient who would more easily tolerate operative assistance is unclear. Regardless, the use of epidural analgesia is now being sought more frequently by patients regardless of the opinion of delivering professional, and is not regarded as failure of labor but as an available option for labor pain relief. Certainly the quality of analgesia with this technique has yet to be equaled by any supportive maneuver or pharmaceutical agent to date. Sincerely yours, Alison Macarthur MD, FRCPC
|
|||
|
|
|||
|
W M Clow, Consultant O&G Withybush Hospital, Haverfordwest, Pembrokeshire
Send response to journal:
|
Sir Mires and his co-authors (1) are to be applauded for the first randomised controlled trial of CTG versus Doppler auscultation of heart rate at admission in labour in low risk obstetric population. By subtracting the study group figures from the whole group analysis it is possible to do an outcome analysis on the patients excluded by virtue of a change in risk. These are shown in Table1.
Table 1 Comparison of neonatal and obstetric outcomes in women who
were excluded from analysis by virtue of increased risk in labour.
Outcome No in
cardiotography
group No in
Doppler group Odds ratio (95% CI)
Cord arterial blood
metabolic acidosis 93/494 108/518 0.90 (0.67 – 1.22)
Apgar score <7
at 5 minutes 11/677 16/697 0.71 (0.29 – 1.64)
Need for IPPV at
resuscitation 6/671 10/700 0.63 (0.19 – 1.91)
Admission to
neonatal intensive care 43/679 60/703 0.74 (0.48 – 1.13)
Hypoxic ischaemic
encephalopathy 2/39 10/56 0.29 (0.03 – 1.30)
Continuous foetal
heart rate monitoring
in labour 574/679 577/704 1.03 (0.88 – 1.21)
Artificial rupture
of membranes 452/679 417/704 1.06 (0.89 – 1.26)
Augmentation of labour 468/679 490/703 0.99 (0.84 – 1.17)
Monitoring of foetal
scalp pH 101/680 101/704 1.04 (0.76 – 1.41)
Epidural anaesthesia
in labour 292/680 304/704 0.99 (0.82 – 1.21)
Operative delivery 289/680 304/704 0.98 (0.81 – 1.20)
Caesarean Section 132/680 122/704 1.12 (0.85 – 1.46)
Since these patients were managed by protocol of their complication rather than the trial randomisation there should be little difference in any outcomes. In this group, however, there is more than a threefold increase in hypoxic ischaemic encephalopathy and a 50% increase in the admission rate to the neonatal intensive care unit. The combination of these worrying trends in neonatal problems reaches statistical significance (Pearson Chi-square test c2 = 5.16, DF=1,CI 95% = 3.84, p<0.05). Since the commonest reason for transfer to this category was induction of labour, and cardiotocography is part of most hospital protocols prior to induction of labour, this significance probably assumes even greater importance. It is in keeping with the known increased risk of neonatal convulsions (2). The findings suggest that the “spirit” of the trial was extended to this “excluded” group with possible adverse effects. The ethos of “low- tech” management, that is found in many obstetric patients and their midwives, may thus be potentially dangerous to the foetus when the risk situation changes even to a mild degree. This paper also illustrates how more than 35% of perceived low risk patients can easily become relatively high risk by the time of labour. Non-interventionist observers will latch on to the1.5% (non-significant) difference in the caesarean section (2% in the above deduced study) and 5.5% increased operative delivery, as a reason to abandon the admission screening test. I am in agreement with the views of Murphy (3) and Grant (4) regarding the possible causal relationship of increased continuous monitoring. However I was also struck by the very close correlation between epidural and operative delivery numbers, notwithstanding the relatively loose relationship described by Morton(5). Most obstetricians will be prepared to balance an increase in caesarean section rate against a substantial reduction in hypoxic ischaemic encephalopathy and intensive care admission rates in those patients where a risk factor may be overlooked or occur unexpectedly. This further analysis of the "concealed" effect of a seemingly sensible change in policy appears to have unexpectedly worrying effects on the management of patients who should not be affected by the change. Does the change alter the level of suspicion in the entire unit? Yours W.M. Clow, Consultant Obstetrician and Gynaecologist, Withybush General Hospital, Haverfordwest, Pembrokeshire SA61 2PZ 1) Mires G, Williams F, Howie P et al,Randomised controlled trial of cardiotocography versus Doppler auscultation of fetal heart at admission in labour in low risk obstetric population BMJ 2001; 322: 1457-1462 2) Thacker SB, Stroup D, Chang M. Continuous electronic heart rate monitoring for fetal assessment during labor (Cochrane Review). In: The Cochrane Library, Issue 2, 2001. Oxford: Update Software. 3) Murphy DJ, Importance of controlling for potential confounding. eBMJ Rapid responses, June 2001 4)Grant S, Statistical or clinical significance? eBMJ Rapid responses, June 2001 5) Morton S C, Williams M S, Keeler E B,et al. Effect of epidural analgesia for labor on the cesarean delivery rate. Obstetrics and Gynecology 1994; 83(6): 1045-1052. Acknowledgement I am very grateful to Mr Philip Jones, research and audit facilitator at Withybush Hospital for the statistical analysis. Conflicting interests: none |
|||
|
|
|||
|
Shane Higgins, Fellow in Maternal-Fetal Medicine Ruyal Women's Hospital, Melbourne,Australia
Send response to journal:
|
Sir, I have the following queries in regard to the Mires paper on admission CTG 1. How did you derive the denominator in the hypoxic ischaemic encephalopathy groups in both Tables 1 and 3. 2. How do you explain the extraordinarily high incidence of metabolic acidosis in the population who remained low risk throughout labour, 18.1% in the admission CTG group and 17.9% in the intermittant auscultation group. 3. Were the 26.6% of babies in the confirmed low risk group who did not have cord arterial pH and base deficit undertaken,omitted because they were so vigorous it was assumed they would not be acidotic or was it incomplete data collection. If it were the latter how can we assume a significant difference in primary outcome was not missed. |
|||
|
|
|||
|
Jane Thomas, Director RCOG Clinical Effectiveness Support Unit RCOG
Send response to journal:
|
RCOG CESU has recently developed the NICE guideline on The Use of Electronic Fetal Monitoring1. One of the potential benefits of guidelines is to highlight gaps in research evidence. As part of the guideline development process we conducted a survey of maternity units in England and Wales to evaluate the current use of EFM in intrapartum care2. The guideline recommends that in women who are healthy and have an uncomplicated pregnancy intermittent auscultation is a suitable method of monitoring during labour. This recommendation was based on the evidence from nine RCTs that have enrolled 18,000 women 3-11and are included in three systematic reviews12-14. This conclusion was reached because screening with continuous EFM should not be conducted in the absence of evidence of benefit. This recommendation will perhaps have a minimal impact in practice as the survey found that only 2.5% of units report using continuous electronic fetal monitoring for all women in labour. However, the majority of maternity units (79%) reported using an admission CTG in the initial assessment of women who are healthy and have had an otherwise uncomplicated pregnancy. When evaluating the research that exists to support the use of admission tests in low risk populations we were surprised to find only a single cohort study that addressed this issue, this cohort study included only 1041 women and related admission tests to fetal distress15. The trial by Mires et al16 is important as it is the first published randomised controlled trial to address this clinically important question, the study is also nearly four times the sample size of previous studies, as such it would have been important to publish this RCT irrespective of whether the results were positive, negative or equivocal. The results of the Mire’s Trial are in line with those of the earlier cohort study and do not change the recommendation in the guideline. We are aware of unease in some quarters that the guideline did not recommended admission CTG and that this recommendation does not concur with current practice in many maternity centres. We hope that this would act as an incentive to researchers and clinicians to conduct and enrol patients in RCTs to answer this important question. Reference List 1. Clinical Effectiveness Support Unit. The use of electronic fetal monitoring. The use and interpretation of cardiotocography in intrapartum fetal surveillance. 2001. London, Royal College of Obstetricians and Gynaecologists. 2. Maternal and Child Health Research Consortium. Confidential Enquiry into Stillbirths and Deaths in Infancy: eighth annual report, due to be published September 2001. 7, 1-113. 2001. London. 3. Vintzileos, A. M., Antsaklis, A., Varvarigos, I., Papas, C., Sofatzis, I., and Montgomery, J. T. A randomized trial of intrapartum electronic fetal heart rate monitoring versus intermittent auscultation [see comments]. Obstet Gynecol 81(6), 899-907. 1993. United States. 4. MacDonald, D., Grant, A., Sheridan-Pereira, M., Boylan, P., and Chalmers, I. The Dublin randomized controlled trial of intrapartum fetal heart rate monitoring. Am J Obstet Gynecol 152(5), 524-539. 1-7-1985. United States. 5. Luthy, D. A., Shy, K. K., van Belle, G., Larson, E. B., Hughes, J. P., Benedetti, T. J., Brown, Z. A., Effer, S., King, J. F., and Stenchever, M. A. A randomized trial of electronic fetal monitoring in preterm labor. Obstet Gynecol 69(5), 687-695. 1987. United States. 6. Haverkamp, A. D., Orleans, M., Langendoerfer, S., McFee, J., Murphy, J., and Thompson, H. E. A controlled trial of the differential effects of intrapartum fetal monitoring. Am J Obstet Gynecol 134(4), 399-412. 15-6- 1979. United States. 7. Kelso, I. M., Parsons, R. J., Lawrence, G. F., Arora, S. S., Edmonds, D. K., and Cooke, I. D. An assessment of continuous fetal heart rate monitoring in labor. A randomized trial. Am J Obstet Gynecol 131(5), 526-532. 1-7-1978. United States. 8. Haverkamp, A. D., Thompson, H. E., McFee, J. G., and Cetrulo, C. The evaluation of continuous fetal heart rate monitoring in high-risk pregnancy. Am J Obstet Gynecol 125(3), 310-320. 1-6-1976. United States. 9. Neldam, S., Osler, M., Hansen, P. K., Nim, J., Smith, S. F., and Hertel, J. Intrapartum fetal heart rate monitoring in a combined low- and high-risk population: a controlled clinical trial. Eur J Obstet Gynecol Reprod Biol 23(1-2), 1-11. 1986. Netherlands. 10. Renou, P., Chang, A., Anderson, I., and Wood, C. Controlled trial of fetal intensive care. Am J Obstet Gynecol 126(4), 470-476. 15-10-1976. United States. 11. Wood, C., Renou, P., Oats, J., Farrell, E., Beischer, N., and Anderson, I. A controlled trial of fetal heart rate monitoring in a low- risk obstetric population. Am J Obstet Gynecol 141(5), 527-534. 1-11-1981. United States. 12. Thacker, S. B. and Stroup, D. F. Continuous electronic heart rate monitoring for fetal assessment during labor. The Cochrane Library (1). 2000. Oxford, Update Software. 13. Vintzileos, A. M., Nochimson, D. J., Guzman, E. R., Knuppel, R. A., Lake, M., and Schifrin, B. S. Intrapartum electronic fetal heart rate monitoring versus intermittent auscultation: a meta-analysis [see comments]. Obstet Gynecol 85(1), 149-155. 1995. United States. 14. Grant A. Monitoring the fetus during labour. In Chalmers I, Enkin M, Keirse MJ, eds. Effective care in pregnancy and childbirth, pp 846-82. Oxford University Press, 1989. 15. Ingemarsson, I., Arulkumaran, S., Ingemarsson, E., Tambyraja, R. L., and Ratnam, S. S. Admission test: a screening test for fetal distress in labor. Obstet Gynecol 68(6), 800-806. 1986. United States. 16. Mires G, Williams F, Howie P. Randomised controlled trial of cardiotocography versus Doppler auscultation of fetal heart at admission in labour in low risk obstetric population. BMJ 2001;322:1457-62. Authors: 1. Ms. Jane Thomas, Director, RCOG Clinical Effectiveness Support Unit (CESU) 2. Dr. Shantini Paranjothy, Clinical Research Fellow, RCOG CESU 3. Mr. Tony Kelly, Clinical Research Fellow, RCOG CESU 4. Ms. Josephine Kavanagh, Research Fellow, RCOG CESU |
|||
|
|
|||
|
Margaret Sheridan-Pereira, Director of Paediatrics; and Director, Health Services Research Unit, Aberdeen Coombe Women's Hospital, Dublin and Health Services Research Unit, University of Aberdeen, Adrian Grant
Send response to journal:
|
We find it surprising that the hypothesis tested by Mires et al (1) did not include a predicted effect on abnormal neonatal neurological signs. Randomised controlled trials of continuous electronic fetal heart rate monitoring during labour have shown a striking effect on neonatal seizures (2), but suggest only a very modest effect on the frequency of low umbilical cord blood pH (3). Mires et al. tabulate data on the frequency of 'hypoxic-ischaemic encephalopathy' and their few data are not incompatible with the reductions seen in previous trials. However, they made no mention of these data in the text of their article. It would be important to know their defining criteria for 'hypoxic-ischaemic encephalopathy', whether any of the babies experienced early seizures, and how they were selected for assessment. There have been comments on the statistical power of this study to detect effects on the primary outcome - low umbilical blood pH. Although this outcome may be more frequent than abnormal neurological signs the evidence cited above suggests that it is less likely to discriminate between the effects of the two fetal monitoring policies compared. Based on data collected in the largest of the previous intrapartum fetal monitoring trials, we believe that the identification of specific neonatal neurological abnormalities (in addition to neonatal seizures) may provide a more frequently occurring, sensitive outcome than neonatal seizures alone In a consecutive series of 1415 liveborn babies in the trial conducted in Dublin (Macdonald et al. 1985), we successfully analysed umbilical venous cord pH in 1378 (97.4%) babies. We applied a standardised neurological assessment to 115 babies with an umbilical pH < 7.2 and a random sample of 96 babies with pH >7.2. Distal hypertonia, abnormal movements and/or abnormal neurobehaviour were observed in 18 (15%) of the acidotic group and only 2 (2%) of the random controls. These symptoms are qualitatively similar to the premonitory signs observed in newborns who convulse within 48 hours of birth. If carefully sought, these symptoms may identify babies who have experienced intrapartum asphyxia of a degree insufficient to cause clinical seizures. Margaret Sheridan-Pereira
Adrian Grant
References 1. Mires G, Williams F, Howie P. Randomised controlled trial of cardiotocography versus Doppler auscultation of fetal heart at admission in labour in low risk obstetric population. BMJ 2001;322:1457-62. 2. Thacker SB, Stroup D, Chang M. Continuous electronic fetal heart rate monitoring for fetal assessment during labor (Cochrane Review). In: The Cochrane Library, Issue 2, 2001. Oxford: Update Software. 3. MacDonald D, Grant A, Sheridan-Pereira M, Boylan P, Chalmers I. The Dublin randomised controlled trial of intrapartum fetal heart rate monitoring. Am J Obstet Gynecol 1985;152:524-39. |
|||
|
|
|||
|
Paola Albertazzi, Clinical lecturer Centre for Metabolic Bone Disease
Send response to journal:
|
Timely recruitment of a sufficient number of volunteers is the pivot to the success of any study. This has recently been the topic of intense debate . Researchers are often faced with the costly dilemma of the number of volunteers that need to be screened in order to obtain sufficient recruits with the characteristics required. In research involving bone density eligibility criteria the quest is often hard. In this short paper we are reporting how the successful recruitment of 153 women in one Centre over 7 months was achieved for a study on the prevention of osteoporosis. The strategy used to estimate the number of subjects needed to be screened and the inexpensive and effective means of achieving the target number of eligible volunteers are indicated. Participants, methods and results Between May and November 2000, 442 women over 60 years old were screened for a double blind placebo controlled clinical trial in osteoporosis prevention. Osteopenia, that is bone density between 1 and 2.5 standard deviations below the young mean, was the main inclusion criteria. Table 1 gives the prevalence of normal bone density osteopenia and osteoporosis in the unselected female population of Hull in comparison with that found in the trial population. Choosing women aged over 60 years both targeted the population at risk of the disease and maximised chances of finding those with osteopenia. Furthermore, osteopenia at either hip or spine was accepted as entry criteria rather than at one site alone. The method of approaching potential volunteers was also felt to be crucial. From past experience media advertising had not proven to be cost effective. General practitioners were also felt to be a poor source of recruitment as, understandably, it is difficult for them to allocate time to clinical trial recruitment in a busy practice. For this trial a group strategy was employed. Local women's groups were approached and a recruitment officer was invited to take part in their meetings and explain the trial. With this method a list of over 500 eligible patients was produced and a screening schedule of an average of 63 patients per month was achieved. Comment We believe that there are two keys to a successful recruitment strategy. The first one is to have a clear indication of the incidence of the condition under investigation in the target population. The second one is to be able to access a sufficient number of motivated volunteers in the allotted time. In our case, a trial recruitment officer was employed who helped by promoting the trial at local women's group meetings. This approach was felt to be the linchpin of our successful recruitment strategy. Group meetings are the ideal means to inform potential volunteers about clinical trials since this gives them a chance to voice and share concerns. Furthermore, when volunteers are evaluating the consequences of the disease versus the possible ineffectiveness of the treatment under trial or the placebo, they seem to be influenced by the decisions of those around them. If they are uncertain as to what action to take they may be taking their 'cues' from the other people in the group. This is not conforming to group pressure but using others behavior as a source of information. The social exchange theory further explains the willingness to volunteer and adhere to clinical trials. The notion that 'we should do unto others as they do unto us' is enhanced in group dynamics and helps mitigate fears of potential lack of direct benefits to the individual by enhancing altruistic feelings of investment in future welfare of the community. Table 1: prevalence of normal bone density osteopenia and osteoporosis in the unselected female local population and in the trial population. Figures are mean ± SD or number and (percentages). Study group Unselected local population Age (years) 69 ±6 50-55 60-69 70-76 Number of subjects 442 6262 306 230 WHO definition Osteoporotic at either hip or spine 80 (18) 446 (7) 58 (19) 69 (30) Osteopenic at either hip or spine 220 (49) 2498 (40) 153 (50) 103 (45) Normal at both sites 142 (32.2) 3323 (53) 95 (31) 58 (25) |
|||