Computerized analysis of normal fetal heart rate pattern
throughout gestation
V. SERRA*†, J. BELLVER*, M. MOULDEN‡ and C. W. G. REDMAN‡
*Unidad de Medicina Materno-Fetal, Instituto Valenciano de Infertilidad and †Departamento de Pediatr´ıa, Obstetricia y Ginecolog´ıa, Facultad de Medicina, Universidad de Valencia, Valencia, Spain and ‡Nuffield Department of Obstetrics and Gynaecology, John Radcliffe Maternity Hospital, University of Oxford, Headington, Oxford, UK
K E Y W O R D S: cardiotocography; computerized cardiotocography; fetal heart rate; fetal monitoring; gestational age
A B S T R A C T
Objective To analyze the evolution of computerized
car-diotocography (cCTG) parameters throughout gestation in a large archive of traces from healthy fetuses.
Methods This was a cross-sectional study of the first
cCTG record from 4412 singleton fetuses with good pregnancy outcome. Normal ranges of cCTG parameters for 25 to 42 weeks were derived from analysis of only one cCTG record per fetus, and the relationship between the parameters and gestational age was investigated.
Results Fetal heart rate (FHR) accelerations, short- and
long-term variation overall, duration of episodes of high and low variation and variation in high episodes increased with advancing gestation. In contrast, maternal perception of fetal movements, basal FHR, variation in low episodes and the time until criteria for normality were met decreased with advancing gestation. Gestational age-related changes in FHR variation were less evident at the lowest percentiles. Episodes of high FHR variation were detected in most fetuses, even at 25 weeks. Opposite trends of basal FHR and variation were observed at 42 weeks. Large decelerations and the frequency and duration of low episodes were also higher at 42 weeks.
Conclusions The characteristics of the normal FHR
pattern are quite defined from early on in gestation, follow a continuous trend with advancing gestation and change abruptly at 42 weeks. Gestational age-related changes are less obvious at the lowest percentiles. Copyright 2009 ISUOG. Published by John Wiley & Sons, Ltd.
I N T R O D U C T I O N
Conventional cardiotocography is probably the most widespread method of fetal surveillance worldwide, but
visual interpretation of the traces is subjective and has unacceptably high intra- and interobserver variability1,2.
These problems are eliminated when using computerized cardiotocography (cCTG), characterized by its objectivity and consistency3. The most validated cCTG system has been under development in Oxford since 1978 and is currently known as the Oxford FetalCare system4. This
analysis concerns data derived from this system.
The magnitude of gestational age-related changes of various cCTG parameters has not been fully investigated in the healthy fetus. Previous reports on the distribution of various cCTG parameters throughout gestation were based on the study of normal traces defined exclusively according to cCTG criteria for normality3,5 – 7, or were
derived from the whole database of cCTG records, irrespective of fetal outcome1; alternatively normal outcome was defined solely on the basis of Apgar scores ≥ 78. The design of the previous studies has
been cross-sectional except for one series comprising 29 uncomplicated pregnancies followed up longitudinally throughout gestation9,10. In addition, neither of these
studies analyzed the changes of all cCTG parameters throughout gestation. Therefore, a large study describing all normative cCTG data defined exclusively on the basis of a comprehensively documented normal fetal outcome was still lacking.
Fetal heart rate (FHR) variation has been shown to be the most useful cCTG indicator of fetal well-being antepartum3,11 – 14. Two normal sources of FHR variation
have been recognized: gestational maturity and episodic changes in fetal behavioral states. The cCTG system takes both issues into consideration. The criteria for normality require detection of at least one episode of high FHR variation (equivalent to active fetal sleep cycles)3,4. The
lower limit of normality of FHR variation that defines such episodes is adjusted for gestational age3,4.
Correspondence to: Prof. V. Serra, Unidad de Medicina Materno-Fetal, Instituto Valenciano de Infertilidad (IVI), Universidad de Valencia, Plaza de la Polic´ıa Local 3, Valencia 46015, Spain (e-mail: [email protected])
As all cCTG parameters are measured numerically, this system is ideal for analyzing objectively the changes throughout gestation. The aim of the present study was to create gestational nomograms of the various objective FHR parameters in a large population of fetuses selected only on the basis of normal outcomes.
M E T H O D S
Antenatal cCTG records were collected in the John Radcliffe Maternity Hospital (Oxford, UK) using the Sonicaid System 8002 (Oxford Instruments Medical Ltd, Surrey, UK) or Apricot Xen personal computers (Apricot, Halesowen, UK) interfaced to FHR monitors (Hewlett-Packard 8040A, Boblingen, Germany; or Sonicaid FM7, Oxford Instruments Medical Ltd). From a large archive of cCTG records (n= 30 897) taken for clinical indications or routine antenatal care in singleton fetuses, we selected the first cCTG trace performed in 4412 fetuses with documented normal outcome (as defined below). This dataset included only one set of results per fetus. Only records of acceptable quality (signal loss <50%) were included.
The cCTG records were divided by gestational age: 25 weeks (n= 48), 26 weeks (n = 60), 27 weeks (n = 83), 28 weeks (n= 110), 29 weeks (n = 110), 30 weeks (n= 138), 31 weeks (n = 142), 32 weeks (n = 192), 33 weeks (n= 198), 34 weeks (n = 206), 35 weeks (n = 246), 36 weeks (n= 377), 37 weeks (n = 523), 38 weeks (n= 557), 39 weeks (n = 520), 40 weeks (n = 520), 41 weeks (n= 324) and 42 weeks (n = 58).
In our cCTG system, the fetal pulse intervals were averaged over 1/16-min epochs (3.75 s) and the baseline fitted using methods previously described3,4. Various
measurements were produced by this computer analysis and defined as follows:
• Basal FHR: baseline heart rate.
• Long-term FHR variation (LTV) overall: minute-by-minute range of the pulse intervals15.
• Short-term FHR variation (STV) overall: sequential epoch-to-epoch variation11. Both the LTV overall and
the STV overall are measured in ms. For an FHR of 140 beats per minute (bpm), an STV of 3.0 ms is equivalent to 1.0 bpm.
• Episodes of high FHR variation (corresponding to active fetal sleep cycles): episodes in which the LTV in at least 5 out of 6 consecutive min was equal to or greater than 32 ms and above the 1st centile for
gestational age3,4. The detection of an episode of high
FHR variation is accepted as evidence of normality and corresponds to the concept of reactivity used in visual analysis3.
• Episodes of low FHR variation (corresponding to quiet fetal sleep cycles): episodes in which the LTV in at least 5 out of 6 consecutive min was less than or equal to 30 ms3,4.
• Accelerations: increases in FHR above the baseline that lasted for longer than 15 s and had an amplitude greater than 10 bpm.
• Decelerations: when a deceleration was encountered, its area was measured in lost beats. A large deceleration was defined when its area was greater than 20 lost beats. Only large decelerations were considered to be of potential clinical importance4.
• Duration of the cCTG record: trace length until criteria for normality met or otherwise extended up to a maximum of 60 min. The first computerized analysis was performed after 10 min of recording and repeated thereafter every 2 min until the Dawes and Redman criteria for normality were met or otherwise continued for up to 60 min3,4. In this study, FHR records were
included whether or not criteria of normality were met provided that the clinical outcome was normal as defined.
Information about each pregnancy and its outcome was extracted from the computerized clinical data system of the John Radcliffe Maternity Hospital. Normal pregnancy outcome was defined as: vaginal delivery or elective Cesarean section before labor at or beyond 37 weeks, neonatal weight above the 10thpercentile for gestational
age and sex, absence of structural or chromosomal abnormalities, normal umbilical artery (UA) pH and base deficit (BD) values at delivery. After vaginal delivery or Cesarean section in labor the criteria were: UA pH > 7.12 and UA BD < 12 mmol/L16. After Cesarean
section before labor they were: UA pH > 7.20 and UA BD < 8 mmol/L. If umbilical cord blood gases were not available (n= 1660; 37.6%) the Apgar scores had to be greater than three at 1 min and seven at 5 min17. Other criteria were lack of admission to the neonatal intensive care unit and intact survival beyond 28 days of postnatal life. Gestational age was checked by ultrasonography at 8–14 weeks in all patients. As usual, the last menstrual period-based date was only corrected if the interval between the ultrasound dates and the menstrual dates was greater than 1 week. The percentile distribution of fetal weight by gestational age and sex was derived from the data of a previous analysis of a population of neonates born in Oxford18. Umbilical arterial and venous blood
gases and pH values were measured (1312 Blood Gas Manager, Instrumentation Laboratories, Milan, Italy) in samples from the double-clamped cord.
Statistical analysis was performed using version 12.0 of the SPSS for Windows statistical package (SPSS Inc., Chicago, IL, USA). Categorical data were expressed as n (%) and compared using the Chi-square test. Continuous data were subjected to the Kolgomorov–Smirnov test for normality. As most cCTG parameters were not normally distributed, continuous data were expressed as the median (interquartile range) and compared using non-parametric tests (the Mann–Whitney U-test and the Kruskal–Wallis test, as appropriate). Analysis of gestational age-related changes was performed comparing the following groups:≤ 32 weeks (n = 883); 33–36 weeks (n = 1027); 37–40 weeks (n= 2120); and 41–42 weeks (n = 382). The following percentiles of the FHR distribution were also calculated: 1st, 3rd, 5th, 10th, 50th, 90th and 95th.
165 160 155 150 145 140 135 130 125 120 115 110 105
Fetal heart rate (bpm)
25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 Gestational age (weeks)
Figure 1 Nomogram of basal fetal heart rate throughout gestation
(n= 4412), indicating 1st( ), 3rd( ), 5th( ), 10th( ), 50th( ), 90th( ) and 95th( ) centiles. 95 90 85 80 75 70 65 60 55 50 45 40 35 30 25 20 15 10
Fetal heart rate variation (ms)
25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 Gestational age (weeks)
Figure 2 Nomogram of long-term fetal heart rate variation overall throughout gestation (n= 4412), indicating 1st( ), 3rd( ), 5th( ), 10th( ), 50th( ), 90th( ) and 95th( ) centiles. 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2
Fetal heart rate variation (ms)
25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 Gestational age (weeks)
Figure 3 Nomogram of short-term fetal heart rate variation overall throughout gestation (n= 4412), indicating 1st( ), 3rd( ), 5th( ), 10th( ), 50th( ), 90th( ) and 95th( ) centiles.
R E S U L T S
Healthy fetuses exhibited decreasing basal FHR (Figure 1) and increasing LTV and STV overall (Figures 2 and 3) from 25 to 41 weeks’ gestation. Opposite trends were unexpectedly found at 42 weeks. The above trends were less obvious for the lowest percentiles of LTV and STV overall, particularly the 1stpercentile, where values
remained fairly constant (or varied within a very narrow range) throughout most of the pregnancy. Episodes of high and low FHR variation were also analyzed separately. LTV in high episodes showed the same increasing trend throughout gestation, with a mild decrease at 42 weeks and flattening at the lowest percentiles, as described above (Figure 4). In contrast, LTV in low episodes decreased gradually throughout gestation, even at the lowest percentile (Figure 5).
Analysis of other cCTG parameters also revealed mild gestational age-related changes (Table 1). The median
36 34 32 30 28 26 24 22 20 18 16 14 12 10 8
Fetal heart rate variation (bpm)
25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 Gestational age (weeks)
Figure 4 Nomogram of long-term variation in episodes of high fetal heart rate variation (equivalent to active fetal sleep cycles)
(n= 4262), indicating 1st( ), 3rd( ), 5th( ), 10th( ), 50th( ), 90th( ) and 95th( ) centiles. 13 12 11 10 9 8 7 6 5 4 3
Fetal heart rate variation (bpm)
25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 Gestational age (weeks)
Figure 5 Nomogram of long-term variation in episodes of low fetal heart rate variation (equivalent to quiet fetal sleep cycles)
(n= 1388), indicating 1st( ), 3rd( ), 5th( ), 10th( ), 50th( ),
Table 1 Changes in computerized fetal heart rate (FHR) parameters across gestation (n= 4412) Gestational age (weeks)
Parameter ≤ 32(n= 883) 33–36 (n= 1027) 37–40 (n= 2120) 41–42 (n= 382) P* Fetal movements per h† 36.0 (15.0–70.0) 35.0 (15.0–65.0) 30.0 (12.0–60.0) 26.0 (11.0–55.0) <0.001
Accelerations 3.0 (1.0–4.0) 4.0 (2.0–6.0) 4.0 (3.0–6.0) 4.0 (3.0–6.3) <0.001
Basal FHR (bpm) 141.0 (135.0–147.0) 139.0 (132.0–145.0) 137.0 (130.0–144.0) 135.0 (128.0–142.0) <0.001 STV overall (ms) 7.7 (6.4–9.4) 8.5 (7.0–10.5) 8.8 (7.0–11.1) 8.9 (7.0–10.9) <0.001 LTV overall (ms) 40.8 (34.6–49.6) 46.4 (37.2–57.5) 47.9 (37.9–60.1) 47.6 (37.5–60.8) <0.001
Records with High episodes 831 (94.1) 998 (97.2) 2060 (97.2) 373 (97.6) <0.001
Minutes of High‡ 7.0 (5.0–10.0) 7.0 (5.0–10.0) 7.5 (6.0–10.0) 8.0 (5.0–11.0) <0.001 LTV in High (bpm)‡ 16.6 (14.6–19.3) 18.9 (15.8–22.3) 19.7 (16.8–23.3) 19.6 (16.4–23.7) <0.001
Records with Low episodes 284 (32.2) 298 (29.0) 668 (31.5) 138 (36.1) NS
Minutes of Low§ 9.0 (6.0–18.0) 9.0 (6.0–17.0) 13.0 (7.0–20.8) 11.5 (7.0–20.0) <0.001 LTV in Low (bpm)§ 8.0 (7.2–9.0) 7.3 (6.6–8.4) 6.8 (5.8–7.9) 6.5 (5.8–7.6) <0.001 Records with large –>20lost
beats – decelerations
14 (1.6) 6 (0.6) 23 (1.1) 9 (2.4) <0.03
Trace length (min)¶ 14.0 (10.0–24.0) 12.0 (10.0–20.0) 12.0 (10.0–22.0) 12.0 (10.0–22.0) <0.001
Results are expressed as median (interquartile range) orn(%).*Kruskal–Wallis test or Chi-square analysis, as appropriate.†As perceived by the mother.‡Calculated only from traces with episodes of high FHR variation.§Calculated only from traces with episodes of low FHR variation.¶Until criteria met or otherwise 60 min. bpm, beats per min; High, episodes of high FHR variation (equivalent to active fetal sleep cycles); Low, episodes of low FHR variation (equivalent to quiet fetal sleep cycles); LTV, long-term FHR variation; NS, non-significant; STV, short-term FHR variation.
number of accelerations increased from 2 at 25–29 weeks to 3 at 30–33 weeks and 4 at 34–42 weeks. The proportion of traces without accelerations decreased from around 8% at 25–28 weeks to 0–1% at 35–41 weeks, but increased to 3.4% at 42 weeks. In the whole dataset, only 1.2% of traces had large decelerations. The proportion remained constantly low (range 0.0–2.3%) throughout gestation, but was higher at 42 weeks (5.2%). The proportion of traces with episodes of high FHR variation and their duration increased throughout ges-tation (Table 1). Most fetuses at the lower end of the gestational age range considered in this study exhibited such episodes, but with a lower proportion: 79.2% at 25 weeks, 90.0% at 26 weeks, 91.6% at 27 weeks and 94.5% at 28 weeks.
The proportion of traces with at least one episode of low variation remained fairly constant within the range of 26–34% between 28 and 41 weeks, being higher at 25 and 42 weeks (43.8% and 53.4%, respectively). In such records, the median number of minutes of low FHR variation increased throughout gestation (Table 1), with these low-variation episodes being longer again at 25 and 42 weeks (15 and 18 min, respectively). Low-variation episodes lasting 50 min or more were detected in 0.6% of the subset of traces with low episodes, and in only 0.2% of all traces.
Maternal perception of fetal movements decreased gradually from nearly 50 movements/h at 25–26 weeks to 21 movements/h at 42 weeks. The median (interquar-tile range) for the whole series was 30.0 (13.0–60.0) movements/h.
Criteria of normality were met in 97.7% of the 4412 traces. The proportion of traces that met criteria of nor-mality increased throughout gestation (P < 0.001), being lower at 25–28 weeks (89.6–93.6%) and at 42 weeks
(94.8%). For all traces, the median duration (interquar-tile range) of the trace until criteria of normality were met (maximum 60 min) was 12.0 (10.0–22.0) min. This duration remained constant throughout gestation within the range of 12.0–14.0 min, being higher at 25 weeks (24.0 min), 26–27 weeks (17.0–18.0 min) and 42 weeks (16.0 min).
D I S C U S S I O N
Most cCTG parameters showed changes with advanc-ing gestation. The effect of gestational age on FHR variation becomes less obvious for the lower percentiles, particularly at the 1stpercentile. Thus, the 1stpercentile
of the STV overall throughout gestation ranged from 3.2 to 4.6 ms in healthy fetuses. This is consistent with the current requirement in the Oxford system’s criteria for normality, namely that when the STV overall is below 4.5 ms then LTV in the episodes of high variation must be above the 3rdpercentile for gestational age4.
The 3rd and 10th percentiles for gestational age of
the LTV in high episodes revealed slightly higher values than previously reported3,4. This is probably because the
previous analyses were of all cCTG records, without selection for pregnancy outcome.
There were several unexpected findings. First, at 42 weeks there were more records with large decelera-tions, more frequent and longer episodes of low variation, increased basal FHR, decreased variation both overall and in high episodes, and fewer fetal movements. These are considered to be less favorable features of clinical health despite the normal outcomes. However it should be noted that all post-term pregnancies were delivered routinely by 42 weeks so their natural evolution was curtailed. Simi-lar findings in post-term pregnancies have been reported before19,20, although changes had not been analyzed from
such a perspective throughout the whole pregnancy, as in the present study. In one of these previous reports19,
a large population of fetuses was studied at 42 weeks (n= 215) and even at ≥ 43 weeks (n = 100). Interest-ingly, the decreasing trend in FHR variation, as well as the other negative cCTG features, were even more evident at≥ 43 weeks. These findings suggest that, indeed, this is the natural evolution in post-term pregnancies. The above changes are difficult to explain and may indicate a poten-tially deleterious effect of postmaturity, although several factors may have contributed. Isolated large decelerations may occur at term, without ominous significance, caus-ing a downward change in FHR variation21. However,
it is unlikely that the observed modifications could be attributed to the presence of such decelerations as these only occurred in 5.2% of the records at 42 weeks.
It was also surprising to find that only a minority of records showed no episodes of high variation in fetuses before 28 weeks of gestation. We previously reported that normal fetuses alternated quiet and active sleep cycles from 28 weeks onwards3, but here we show that episodes
of high variation can also be detected regularly as early as 25 weeks. In fetuses at 25–28 weeks, the proportion of cCTG records without accelerations and episodes of high variation was lower in our study than that reported in a previous small series22.
It is also notable that large decelerations occurred in 1.2% of the records from healthy fetuses at all gestational ages, no accelerations occurred in 1.5% and that the time required to fulfill the criteria for cCTG normality was similar throughout gestation (median duration 12.0 min). The records were taken with the women in a semirecumbent position, so it is unlikely that
the decelerations can be ascribed to episodes of supine hypotension. Such decelerations may be normal features of healthy fetuses in response to transient episodes of cord compression21. Our observation is consistent with
the fact that decelerations are present in, and accelerations absent from, 5–8% of otherwise normal cCTG records with no ominous prognostic significance21,23. The cCTG
record length in the present study is slightly shorter (12.0 min) than the mean duration previously reported in an unselected population of fetuses (15.5 min)24.
This analysis confirms that antenatal episodes of low FHR variation lasting more than 50 min are extremely unlikely (only 0.2% in our whole series) for healthy fetuses as previously reported3,4,25. In addition, we have shown that episodes of low FHR variation last longer, and have decreased variation at term. Despite this, LTV and STV overall increase with advancing gestation owing to increases in the duration and amplitude of the episodes of high variation.
In summary, we provide reference nomograms from 25 to 42 weeks of the various objective FHR parameters based on a large population of healthy fetuses. The characteristics of the normal FHR pattern are quite well defined from early on in gestation, follow a continuous trend with advancing gestation, and change abruptly at 42 weeks. Gestational changes of the FHR variation are less obvious at the lowest percentiles.
A C K N O W L E D G M E N T S
V. Serra was funded in part by a grant from the Spanish Fondo de Investigaciones Sanitarias (FIS 96/0937) and the University of Valencia, Spain; M. Moulden received financial support from Oxford Instruments, Medical Division. This work was in part supported by the NIHR Biomedical Research Centre, Oxford, UK.
R E F E R E N C E S
1. Bernardes J, Costa-Pereira A, Ayres-de-Campos D, van Geijn HP, Pereira-Leite L. Evaluation of interobserver agreement of cardiotocograms. Int J Gynaecol Obstet 1997; 57: 33–37. 2. Figueras F, Albela S, Bonino S, Palacio M, Barrau E,
Her-nandez S, Casellas C, Coll O, Cararach V. Visual analysis of antepartum fetal heart rate tracings: inter- and intra-observer agreement and impact of knowledge of neonatal outcome. J Perinat Med 2005; 33: 241–245.
3. Dawes GS, Moulden M, Redman CWG. Improvements in computerized fetal heart rate analysis antepartum. J Perinat Med 1996; 24: 25–36.
4. Pardey J, Moulden M, Redman CWG. A computer system for the numerical analysis of nonstress tests. Am J Obstet Gynecol 2002; 186: 1095–1103.
5. Dawes GS, Houghton CRS, Redman CWG, Visser GHA. Pat-tern of the normal human fetal heart rate. Br J Obstet Gynaecol 1982; 89: 276–284.
6. Dawes GS, Redman CWG, Smith JH. Improvements in the registration and analysis of fetal heart rate records at the bedside. Br J Obstet Gynaecol 1985; 92: 317–325.
8. Park MI, Hwang JH, Cha KJ, Park YS, Koh SK. Computerized analysis of fetal heart rate parameters by gestational age. Int J Gynaecol Obstet 2001; 74: 157–164.
9. Nijhuis IJM, ten Hof J, Mulder EJH, Nijhius JG, Narayan H, Taylor DJ, Westers P, Visser GHA. Numerical fetal heart rate analysis: nomograms, minimal duration of recording and intrafetal consistency. Prenat Neonat Med 1998; 3: 314–322.
10. Nijhuis IJM, ten Hof J, Mulder EJ, Nijhius JG, Narayan H, Taylor DJ, Visser GHA. Fetal heart rate in relation to its variation in normal and growth retarded fetuses. Eur J Obstet Gynecol Reprod Biol 2000; 89: 27–33.
11. Street P, Dawes GS, Moulden M, Redman CWG. Short-term variation in abnormal antenatal fetal heart rate records. Am J Obstet Gynecol 1991; 165: 515–523.
12. Dawes GS. Computerized analysis of the fetal heart rate. Eur J Obstet Gynecol Reprod Biol 1991; 42: S5–S8.
13. Turan S, Turan OM, Berg C, Moyano D, Bhide A, Bower S, Thilaganathan B, Gembruch U, Nicolaides K, Harman C, Baschat AA. Computerized fetal heart rate analysis, Doppler ultrasound and biophysical profile score in the prediction of acid-base status of growth-restricted fetuses. Ultrasound Obstet Gynecol 2007; 30: 750–756.
14. Serra V, Moulden M, Bellver J, Redman CWG. The value of the short-term fetal heart rate variation for timing the delivery of growth-retarded fetuses. BJOG 2008; 115: 1101–1107.
15. Henson GL, Dawes GS, Redman CWG. Antenatal fetal heart-rate variability in relation to fetal acid-base status at caesarean section. Br J Obstet Gynaecol 1983; 90: 516–521.
16. Sheil O. Microprocessor analysis of the fetal heart rate in labour.
University College Dublin, National University of Ireland, Thesis, 1991.
17. Sykes GS, Molloy PM, Johnson P, Gu W, Ashworth F, Stir-rat GM, Turnbull AC. Do Apgar scores indicate asphyxia? Lancet 1982; 1: 494–496.
18. Yudkin PL, Aboualfa M, Eyre JA, Redman CWG, Wilkinson AR. New birthweight and head circumference centiles for ges-tational ages 24 to 42 weeks. Early Hum Dev 1987; 15: 45–52. 19. Mandruzzato G, Meir YJ, D’Ottavio G, Conoscenti G, Dawes GS. Computerised evaluation of fetal heart rate in post-term fetuses: long term variation. Br J Obstet Gynaecol 1998; 105: 356–359.
20. Bartnicki J, Ratanasiri T, Meyenburg M, Saling E. Postterm pregnancy: computer analysis of the antepartum fetal heart rate patterns. Int J Gynaecol Obstet 1992; 37: 243–246. 21. Dawes GS, Lobb MO, Mandruzzato GP, Moulden M,
Red-man CWG, Wheeler T. Large fetal heart rate decelerations at term associated with changes in fetal heart rate variation. Am J Obstet Gynecol 1993; 168: 105–111.
22. Roberts D, Kumar B, Tincello DG, Walkinshaw SA. Comput-erised antenatal fetal heart rate recordings between 24 and 28 weeks of gestation. BJOG 2001; 108: 858–862.
23. Ribbert LS, Fidler V, Visser GHA. Computer-assisted analysis of normal second trimester fetal heart rate patterns. J Perinat Med 1991; 19: 53–59.
24. Dawes GS, Lobb MO, Moulden M, Redman CWG, Wheeler T. Antenatal cardiotocogram quality and interpretation using computers. Br J Obstet Gynaecol 1992; 99: 791–797. 25. Visser GH, Dawes GS, Redman CWG. Numerical analysis of
