- Email: [email protected]
Patient-recorded domiciliary fetal monitoring
Patient-recorded domiciliary fetal monitoring P. C. Lindsay, MA, MB, BChir, R. Beveridge, Y. Tayob, MA, MRCOG, L. M. Irvine, MRCOG, I. D. Vellacott, MRCOG, J. A. Giles, FRCS, MRCOG, S. Y. Hussain, MRCOG, and P. M. S. O'Brien, MD, MRCOG London, England Domiciliary fetal monitoring via a telephone link has been in clinical use at the Royal Free Hospital for 2 years, during which time 858 traces of 134 patients who were moderately at risk have been transmitted to the hospital. All recognized abnormal heart rate patterns were detected during the study period. In seven patients obstetric intervention took place on the basis of the domiciliary fetal monitoring recording and subsequent hospital monitoring. There were no intrauterine or neonatal deaths, and there was no overall increase in obstetric operative intervention. There appeared to be a significant reduction in the inconvenience suffered by the patient and her family, compared with conventional monitoring. Fetal heart recordings performed solely by the patient were of a high quality and reduced the demand on midwifery resources. To allow this only 3 hours of midwifery time daily was required and no additional demands on obstetric staff were made. It is anticipated that domiciliary fetal monitoring, performed by patients, will become an essential and established service throughout the United Kingdom (and probably Europe) as the principal means of monitoring fetal well-being. (AM J OSSTET GVNECOl1990;162:466-70.)
Key words: Cardiotocography, self-recorded domiciliary telephone transmission
The use of unstressed antenatal cardiotocography has become established as one of the current principal means of assessing fetal well-being. The technique unfortunately entails frequent and time-consuming hospital visits and may even necessitate hospital admission. Additionally, it imposes great demands on midwifery expertise that would otherwise be used elsewhere. Transmission of fetal heart traces by telemetry was developed in the early 1980s at the obstetric units at Cardiff and Cambridge University. These techniques were originally devised for real-time transmission, I but recent developments in microprocessor technology by the Cardiff group and Huntleigh Technology has allowed the recording and subsequent rapid transmission of compressed stored data!' 3 A report on both the technologic development and the clinical and technical evaluation in field trials was published recently.'" We report here the first and largest series to date of the commercially available system in clinical use. Eight hundred fifty-eight traces of 134 women are reported together with the analyses of these traces and the clinical outcomes. All recordings were made by the patients in their homes.
From the Academic Department of Obstetncs and Gynaecology, Royal Free Hospital. Received for publication May 2, 1989; revised August 1, 1989; accepted August 7, 1989. Reprint requests: Professor P. M. S. O'Brien, University of Keele, School of Postgraduate MedicIne and Biologzcal Sciences, Thornburrow Dr., Hartshill, Stoke on Trent, UK ST4 7QB.
611115885
466
Patients and methods
Patients. One hundred thirty-four "at risk" patients were referred for domiciliary monitoring, after 28 weeks of pregnancy. The ages of the patients at delivery ranged from 19 to 43 years with a mean of 28 years. The indications included reduced fetal movements, current or previous intrauterine growth retardation, oligohydramnios, previous stillbirth, antepartum hemorrhage, hypertension, gestational diabetes, macrosomia, postdates, and loss of or failure to gain weight (Table I). Domiciliary fetal monitoring equipment. The technical details of the commercially available Huntleigh domiciliary fetal monitor (Huntleigh Technology PLC, Healthcare Division, Cardiff) have been reported previously.2.3 The equipment consists of remote units (five were used in the present report), a hard-disk PC type microprocessor and printer, an electronic modem, and the use of a dedicated standard British Telecom telephone line. The remote unit, the domiciliary fetal monitoring unit itself, is briefcase-sized. It is powered from both mains and an integral rechargeable unit and has a Doppler wide-beam ultrasonographic transducer, microprocessor, and integral acoustic modem. There is an event marker to record fetal movement and command touch switches are simple to use. All remote unit functions were patient-operated. Procedure. Patients were instructed in the use of domiciliary fetal monitoring by the fetal monitoring unit obstetric nurse-midwife. Each patient was in-
Patient-recorded domiciliary fetal monitoring
Volume 162 Number 2
Be f' tl~
467
.t.
B
19
I:
.J
- -c---:- +
:00
25
Minut~s
-1------t--
-
- --
.
'----'ito -+. - ---
i60
1-
--
-
'. -1-- . -
2(
- -
11
. --.
-t-1- I I
3~~i+~~-\Wi~~~~~~~q-~.~~ ,
. -I-----'·F
- .- --- - .. - ----1-00--- --
-r---=r-h -
-1-
---,-+ -
, -, I
80
- t-- ~'"
-+ -
0--1·
-
1-
- -. -
1- -- J-------
--t- -
----'--- ----,--
f---r-
+ --
f---t- -
rI '
-
-
-
--"1' -' - :-
~
. ----, -
-------,-bo .. -- - - --.- -" - -- -'-- ---+-+ ~ -
,-
: _,;, : -- -+ _-'~?
__ _
-- ~-
-
,- .. -- -
- ;-
1-
-J _
---1-
:--, -~
,- I---+--- -+--
--i- h - 8
--. f-t--H-
Fig. l. Simultaneo us recording of domiciliary fetal monitoring and conventional c ardiotocography obtained from a Hewlett Packard model 8040 A.
structed for 20 minutes in the u se of the remote unit. Recordings were transmitted the next day and thereafter daily or on alternate days as requested by the obstetrician_ The mother located and recorded the fetal heart for 30 minutes, the heart rate data then being converted to a digital code and stored by the microprocessor. The base station was phoned later at a prearranged time (within 1 hour). After consultation with the fetal monitoring obstetric nurse-midwife the telephone handset was placed into the acoustic couple; after being triggered by the base station, the information was transmitted in compressed digital form during approximately 45 seconds. The base station computer used a dedicated program to receive and analyze the signal and display and print the fetal heart trace. The traces were assessed immediately for technical quality and content by the obstetric nurse-midwife , and the patients were given a provisional r eport. All traces were subsequently reviewed by the trained obstetrician on call. Traces of poor technical quality were repeated . Patients whose traces were abnormal were brought to the hospital immediately for conventional cardiotocography. The content of each trace was assessed as for conventional cardiotocography : normal baseline of 120 to 160 beats / min and two accelerations of at least 15 beats/ min for at least 15 seconds with good baseline variability. Fig. 1 shows a normal domiciliary fetal monitoring recording compared with a conventional trace obtained from a Hewlett Packard model 8040A. The acceptance rate is the percentage of time during a recording for which fetal heart ra te data are available. This was automatically computed by the microprocessor, which calculated the percenta ge loss of fetal heart
Table I. Indications for domiciliary fetal monitoring N o. of
No. of
IndicatIOn
patients
traces
Current intrauterine growth retardation Reduced fetal movements Postdates Static or reduced weight Small antepartum hemorrhage Previous stillbirth Mild hypertension Gestational diabetes Macrosomia Maternal anxiety Oligohydramnios Previous intrauterine growth retardation
40
372
29 21
219
II 8
32 32
8
42
TOTAL
7 3 2 2 2
45
44
5 4
21
16
I
26
134
858
rate data not caused by signal loss during transmission but caused by failure of pickup during recording (a frequent occurrence with conventional cardiotocography) . Acceptance rates of ~70 % were considered easily interpretable; recording >40% to 69% was considered interpretable if two accelerations were identified. Results
Patient data. One hundred thirty-four patients were accepted for domiciliary fetal monitoring. Seventy-four were primigravid. The length of gestation at referral ranged from 30 to 42 weeks. The most common indications for monitoring were suspected or actual in-
468
Lindsay et a!.
February 1990 Am J Obstet Gynecol
300
en c
0
.(j)
en
·E
200
en c
-«S L-
0
L-
Q)
.0
E
100
:::J
Z
o
10
20
30
40
50
60
70
80
90
100
Acceptance % Fig. 2. Number of patients achieving acceptance percentages.
Table II. Obstetric intervention resulting from domiciliary fetal monitoring Patient data and monitoring indication
Result of domiciliary fetal monitonng
I. 30 yr old Primiparous 39 wk gestation Reduced fetal movement 2. 28 yr old Primiparous 39 wk gestation Intrauterine growth retardation 3.20 yr old Multiparous Term gestation "Maternal anxiety" 4. 31 yr old Multiparous 38 wk gestation Intrauterine growth retardation 5. 27 yr old Multiparous >41 wk gestation Postmaturity Intrauterine growth retardation 6.38 yr old Multiparous; breech 37 wk gestation Intrauterine growth retardation 7.28 yr old Multiparous; breech Suspected intrauterine growth retardation
Unreactive
Cardiotocography Unreactive
Induced normal labor 3.5 kg 9 and 10
Decelerations
Cardiotocography Further decelerations
2.5 kg 9 and 9
Decelerations
Cardiotocography Further decelerations Scan No growth
Normal delivery 3.4 kg 8 and 9
Poorly reactive x 3
Cardiotocography Unreactive
Normal delivery 2.8 kg 8 and 9
Unreactive
Cardiotocography Not confirmed Scan No growth Cardiocotography Very flat and unreactive
Planned LSCS performed 10 days early. no growth 2.1 kg 3 and 9
Unreactive
Unreactive
LSCS. Low-segment cesarean section.
Action
Outcome (birth weight and Apgar SCOTe at 1 and 5 mm)
Forceps; fetal distress
Vaginal delivery ; fetal distress 2.6 kg 8 and 9
Planned LSCS performed 2 wk early 2.8 kg 7 and 9
Volume 162 Number 2
trauterine growth retardation, reduced fetal movements, postdates, and loss of or failure to gain weight (Table I). A small group who benefited greatly from the reassurance provided were those referred because of a previous stillbirth. Outcome of domiciliary fetal monitoring traces. Acceptance rates were high with more than 643 traces (75%) demonstrating rates in excess of 70% (Fig. 2). In view of adequate content (reactivity), a further 9% of traces were considered interpretable. Mean acceptance rates improved somewhat with increasing gestation: before 32 weeks, 74.8%; 32 to 35 weeks, 75.8%; 36 to 39 weeks, 79.1 %; 40 to 42 weeks, 81.2%. Domiciliary fetal monitoring traces were repeated without admission to the hospital either for poor technical quality (43 or 4.9%) or for poor reactivity, other parameters being normal (58 or 6.8%). After 45 (5.2%) transmissions patients were asked to undergo cardiotocography because of a suspicious trace. These abnormal domiciliary fetal monitoring traces included 19 with an unreactive trace or poor baseline variability, four with a fetal tachycardia, two with bradycardia, and six with probable or definite decelerations. A further 14 had traces of repeated poor technical quality. In 38 of the 45 (84%) the repeat trace was satisfactory, but in the remaining seven obstetric intervention was necessary; these data are summarized in Table II. Five patients underwent induction oflabor as a result of abnormal domiciliary fetal monitoring with subsequent confirmation by cardiotocography. Two patients had their planned elective cesarean sections performed earlier. The obstetric outcome of the monitored group was similar to that of the general unit with a cesarean section rate of 16% and a forceps rate of 15%. The induction rate of28% was greater, being compared with 15%. Sixty-seven percent of the group, however, had birth weights below the 50th percentile, reflecting the recruitment of this moderately at risk grou p. The perinatal mortality rate was zero compared with 9 in 1000 for the same period in this unit. Ten patients initially referred for domiciliary fetal monitoring were eliminated from the data analysis because no trace was ever actually transmitted. Four women with postdate gestations went into labor before the first transmission. Five were incapable of operating the monitor after instruction and hence were not accepted to the monitoring program. One (a schoolteacher), having been instructed, refused to use the device because the hospital would not pay the telephone charges. Comment
Conventional unstressed fetal heart rate monitoring is well established and is the most frequently used tech-
Patient-recorded domiciliary fetal monitoring
469
Table III. Obstetric and neonatal outcome Mode of delivery Spontaneous vaginal delivery Anderson's forceps Kielland's forceps Cesarean section Elective Emergency Neonatal outcome Low Apgar score «6 at 5 min) Transferred to special care baby unit Birth weight percentiles <10th 10th-24th 25th-49th 50th-74th 75th-90th >90th Stillbirths Early neonatal deaths
89 17
4
10 13
2 3
24
26
38
32
9 5
o o
nique of fetal surveillance despite limited supportive research. The results of studies have not demonstrated universal benefit,', 6 although it has been clearly demonstrated that perinatal deaths are generally preceded by abnormal heart rate patterns. 5•S It is impossible to monitor large numbers of pregnancies with adequate frequency to detect reliably these abnormalities. The aim of domiciliary fetal monitoring recording has never been to provide any more than what is already available with conventional cardiotocography. In practical terms it is perhaps a little more difficult to interpret quantitatively the loss of baseline variability. Identification of all other abnormalities was directly comparable with that of conventional cardiotocography. Domiciliary fetal monitoring should be perceived as a screening technique that can be performed frequently but with minimal disruption to the patient and to the functioning of a perinatal unit. In this study we have looked at pregnancies with sufficient risk to require regular monitoring either directly from the clinic or after a short period of evaluation as an antenatal inpatient. High-risk pregnancies where admission would normally be required were not studied, although it may be seen in the future that these too are suitable for such surveillance. The degree of risk had been determined from the history before study entry and was confirmed to an extent by the number of babies below the 50th percentile. In spite of this risk, the overall obstetric outcome was in line with that of the unit as a whole and the fetal outcome at birth was surprisingly good with no perinatal losses. Allowing the patient to monitor her own fetal heart rate from home reduces pressure on ward, outpatient, and community midwives. One would anticipate a reduced acceptance rate. In fact, the converse is true;
470
Lindsay et al.
",hen the acceptance rates of this group of patients are compared with those of the midwife-recorded traces' in the multicenter study, those of the patients appear superior. The microprocessor technique of data storage and compression allows significant advantages over realtime techniques since telephone interference is minimized, telephone costs are reduced, and the fetal monitoring "sister" can be present at the base station for a much shorter duration of the transmission, and not the whole 30 minutes of the fetal heart recording. Although the system has been well tried in challenging situations such as distant islands off the Scottish mainland and the Harlem district of New York, our evaluation of the domiciliary fetal monitoring system has been used to assess its use in moderately "at risk" pregnancies in the Hampstead Health District. It has proved to be highly acceptable to patients, producing reassurance in almost all monitored women and indifference in a few. Acceptability and enthusiasm also have been shown by midwifery and virtually all obstetric staff. The quality of traces and the high acceptance rates appear as good in patient-recorded traces as in those performed by specifically trained midwives.' The fetal and neonatal outcome in this "at risk" group is as good as it is in the total hospital population. This has been achieved without an increase of operative obstetric intervention; the induction rate of 28%, not surprisingly, was higher. Domiciliary fetal monitoring reduces disruption to the lives of patients and their families . It enables "at risk" patients, who should be resting, to remain at home when they might otherwise require time-consuming visits to the hospital. It provides us with the ability to monitor more patients more frequently. We believe that there would be a demonstrable reduction in morbidity, mortality, and inpatient bed occupancy. These factors would release midwifery expertise and improve our overall obstetric service. Much larger numbers of patients will be needed to demonstrate this; these studies are ongoing. Further studies to assess the value of domiciliary fetal monitoring in comparison to the standard nonstress test (twice-
February 1990
Am J Obstet Gynecol
daily fetal movement charts) as sensitive means of monitoring fetal well-being are required. From the experience of this study certain recommendations can be made. It is virtually mandatory that a dedicated telephone line with a specific number is allocated to avoid the inevitable problems with hospital switchboards. Whether the system is run from an antenatal clinic (easier for referral), a labor ward (24-hour availability), or a special unit depends on local needs. The person managing the base station should be an obstetric nursemidwife available throughout each antenatal clinic. All traces should be reported early the same day to an experienced obstetrician. We thank Huntleigh Technology PLC, particularly D. Stanger and T . Justice. We also thank P. Chalk, S. M. Tuck, P. Walker, and Professor R. W. Shaw and their obstetric patients who took part in this evaluation. REFERENCES 1. Dalton KJ, Dawson AJ, Gough NAJ. Long distance telemetry of fetal heart rate from patients' own home using telephone network. Br Med J 1983;286: 1545. 2. Gough NAL, Dawson AJ, Tomkins TJ. Antenatal fetal heart rate recording and subsequent fast transmission by a distributed microprocessor-based system. Int J Biomed Comp 1986;18:61-5. 3. Dawson AJ. Middlemiss C, Jones EM, Gough NAJ. Fetal heart rate monitoring by telephone. 1. Development of an integrated system in Cardiff. Br J Obstet Gynaecol 1988;95: 10 18-1 03. 4. James D. Peralta B, Porter S, et al. Fetal heart rate monitoring by telephone. II. Clinical experience in four centres with a commercially produced system. Br J Obstet Gynaecol 1988;95: 1024-9. 5. Flynn AM. Kelly J , Mansfield H , Needham P, O C ' onoT M, Viegas O. A randomised controlled trial of non-stress antepartum cardiotography. Br J Obstet Gynaecol 1982; 89:427-33. 6. Lumley J, Lester A, Anderson I, Renou P, Wood C. A randomised trial of weekly cardiotocography in high risk obstetric patients. Br J Obstet Gynaecol 1983 ;90:1016-26. 7. Flynn AM, Kelly J. Evaluation of fetal well-being by antepartum fetal heart rate monitoring. BT Med J 1977; I : 936-9. 8. Rochard F, Schifrin BS, Goupil F, Legrand H, Blottiere J. Sureau J. Nonstressed fetal heart rate monitoring in the antepartum period . AM J OBSTET GYNECOL 1976; 126:699706.
Key words: Cardiotocography, self-recorded domiciliary telephone transmission
The use of unstressed antenatal cardiotocography has become established as one of the current principal means of assessing fetal well-being. The technique unfortunately entails frequent and time-consuming hospital visits and may even necessitate hospital admission. Additionally, it imposes great demands on midwifery expertise that would otherwise be used elsewhere. Transmission of fetal heart traces by telemetry was developed in the early 1980s at the obstetric units at Cardiff and Cambridge University. These techniques were originally devised for real-time transmission, I but recent developments in microprocessor technology by the Cardiff group and Huntleigh Technology has allowed the recording and subsequent rapid transmission of compressed stored data!' 3 A report on both the technologic development and the clinical and technical evaluation in field trials was published recently.'" We report here the first and largest series to date of the commercially available system in clinical use. Eight hundred fifty-eight traces of 134 women are reported together with the analyses of these traces and the clinical outcomes. All recordings were made by the patients in their homes.
From the Academic Department of Obstetncs and Gynaecology, Royal Free Hospital. Received for publication May 2, 1989; revised August 1, 1989; accepted August 7, 1989. Reprint requests: Professor P. M. S. O'Brien, University of Keele, School of Postgraduate MedicIne and Biologzcal Sciences, Thornburrow Dr., Hartshill, Stoke on Trent, UK ST4 7QB.
611115885
466
Patients and methods
Patients. One hundred thirty-four "at risk" patients were referred for domiciliary monitoring, after 28 weeks of pregnancy. The ages of the patients at delivery ranged from 19 to 43 years with a mean of 28 years. The indications included reduced fetal movements, current or previous intrauterine growth retardation, oligohydramnios, previous stillbirth, antepartum hemorrhage, hypertension, gestational diabetes, macrosomia, postdates, and loss of or failure to gain weight (Table I). Domiciliary fetal monitoring equipment. The technical details of the commercially available Huntleigh domiciliary fetal monitor (Huntleigh Technology PLC, Healthcare Division, Cardiff) have been reported previously.2.3 The equipment consists of remote units (five were used in the present report), a hard-disk PC type microprocessor and printer, an electronic modem, and the use of a dedicated standard British Telecom telephone line. The remote unit, the domiciliary fetal monitoring unit itself, is briefcase-sized. It is powered from both mains and an integral rechargeable unit and has a Doppler wide-beam ultrasonographic transducer, microprocessor, and integral acoustic modem. There is an event marker to record fetal movement and command touch switches are simple to use. All remote unit functions were patient-operated. Procedure. Patients were instructed in the use of domiciliary fetal monitoring by the fetal monitoring unit obstetric nurse-midwife. Each patient was in-
Patient-recorded domiciliary fetal monitoring
Volume 162 Number 2
Be f' tl~
467
.t.
B
19
I:
.J
- -c---:- +
:00
25
Minut~s
-1------t--
-
- --
.
'----'ito -+. - ---
i60
1-
--
-
'. -1-- . -
2(
- -
11
. --.
-t-1- I I
3~~i+~~-\Wi~~~~~~~q-~.~~ ,
. -I-----'·F
- .- --- - .. - ----1-00--- --
-r---=r-h -
-1-
---,-+ -
, -, I
80
- t-- ~'"
-+ -
0--1·
-
1-
- -. -
1- -- J-------
--t- -
----'--- ----,--
f---r-
+ --
f---t- -
rI '
-
-
-
--"1' -' - :-
~
. ----, -
-------,-bo .. -- - - --.- -" - -- -'-- ---+-+ ~ -
,-
: _,;, : -- -+ _-'~?
__ _
-- ~-
-
,- .. -- -
- ;-
1-
-J _
---1-
:--, -~
,- I---+--- -+--
--i- h - 8
--. f-t--H-
Fig. l. Simultaneo us recording of domiciliary fetal monitoring and conventional c ardiotocography obtained from a Hewlett Packard model 8040 A.
structed for 20 minutes in the u se of the remote unit. Recordings were transmitted the next day and thereafter daily or on alternate days as requested by the obstetrician_ The mother located and recorded the fetal heart for 30 minutes, the heart rate data then being converted to a digital code and stored by the microprocessor. The base station was phoned later at a prearranged time (within 1 hour). After consultation with the fetal monitoring obstetric nurse-midwife the telephone handset was placed into the acoustic couple; after being triggered by the base station, the information was transmitted in compressed digital form during approximately 45 seconds. The base station computer used a dedicated program to receive and analyze the signal and display and print the fetal heart trace. The traces were assessed immediately for technical quality and content by the obstetric nurse-midwife , and the patients were given a provisional r eport. All traces were subsequently reviewed by the trained obstetrician on call. Traces of poor technical quality were repeated . Patients whose traces were abnormal were brought to the hospital immediately for conventional cardiotocography. The content of each trace was assessed as for conventional cardiotocography : normal baseline of 120 to 160 beats / min and two accelerations of at least 15 beats/ min for at least 15 seconds with good baseline variability. Fig. 1 shows a normal domiciliary fetal monitoring recording compared with a conventional trace obtained from a Hewlett Packard model 8040A. The acceptance rate is the percentage of time during a recording for which fetal heart ra te data are available. This was automatically computed by the microprocessor, which calculated the percenta ge loss of fetal heart
Table I. Indications for domiciliary fetal monitoring N o. of
No. of
IndicatIOn
patients
traces
Current intrauterine growth retardation Reduced fetal movements Postdates Static or reduced weight Small antepartum hemorrhage Previous stillbirth Mild hypertension Gestational diabetes Macrosomia Maternal anxiety Oligohydramnios Previous intrauterine growth retardation
40
372
29 21
219
II 8
32 32
8
42
TOTAL
7 3 2 2 2
45
44
5 4
21
16
I
26
134
858
rate data not caused by signal loss during transmission but caused by failure of pickup during recording (a frequent occurrence with conventional cardiotocography) . Acceptance rates of ~70 % were considered easily interpretable; recording >40% to 69% was considered interpretable if two accelerations were identified. Results
Patient data. One hundred thirty-four patients were accepted for domiciliary fetal monitoring. Seventy-four were primigravid. The length of gestation at referral ranged from 30 to 42 weeks. The most common indications for monitoring were suspected or actual in-
468
Lindsay et a!.
February 1990 Am J Obstet Gynecol
300
en c
0
.(j)
en
·E
200
en c
-«S L-
0
L-
Q)
.0
E
100
:::J
Z
o
10
20
30
40
50
60
70
80
90
100
Acceptance % Fig. 2. Number of patients achieving acceptance percentages.
Table II. Obstetric intervention resulting from domiciliary fetal monitoring Patient data and monitoring indication
Result of domiciliary fetal monitonng
I. 30 yr old Primiparous 39 wk gestation Reduced fetal movement 2. 28 yr old Primiparous 39 wk gestation Intrauterine growth retardation 3.20 yr old Multiparous Term gestation "Maternal anxiety" 4. 31 yr old Multiparous 38 wk gestation Intrauterine growth retardation 5. 27 yr old Multiparous >41 wk gestation Postmaturity Intrauterine growth retardation 6.38 yr old Multiparous; breech 37 wk gestation Intrauterine growth retardation 7.28 yr old Multiparous; breech Suspected intrauterine growth retardation
Unreactive
Cardiotocography Unreactive
Induced normal labor 3.5 kg 9 and 10
Decelerations
Cardiotocography Further decelerations
2.5 kg 9 and 9
Decelerations
Cardiotocography Further decelerations Scan No growth
Normal delivery 3.4 kg 8 and 9
Poorly reactive x 3
Cardiotocography Unreactive
Normal delivery 2.8 kg 8 and 9
Unreactive
Cardiotocography Not confirmed Scan No growth Cardiocotography Very flat and unreactive
Planned LSCS performed 10 days early. no growth 2.1 kg 3 and 9
Unreactive
Unreactive
LSCS. Low-segment cesarean section.
Action
Outcome (birth weight and Apgar SCOTe at 1 and 5 mm)
Forceps; fetal distress
Vaginal delivery ; fetal distress 2.6 kg 8 and 9
Planned LSCS performed 2 wk early 2.8 kg 7 and 9
Volume 162 Number 2
trauterine growth retardation, reduced fetal movements, postdates, and loss of or failure to gain weight (Table I). A small group who benefited greatly from the reassurance provided were those referred because of a previous stillbirth. Outcome of domiciliary fetal monitoring traces. Acceptance rates were high with more than 643 traces (75%) demonstrating rates in excess of 70% (Fig. 2). In view of adequate content (reactivity), a further 9% of traces were considered interpretable. Mean acceptance rates improved somewhat with increasing gestation: before 32 weeks, 74.8%; 32 to 35 weeks, 75.8%; 36 to 39 weeks, 79.1 %; 40 to 42 weeks, 81.2%. Domiciliary fetal monitoring traces were repeated without admission to the hospital either for poor technical quality (43 or 4.9%) or for poor reactivity, other parameters being normal (58 or 6.8%). After 45 (5.2%) transmissions patients were asked to undergo cardiotocography because of a suspicious trace. These abnormal domiciliary fetal monitoring traces included 19 with an unreactive trace or poor baseline variability, four with a fetal tachycardia, two with bradycardia, and six with probable or definite decelerations. A further 14 had traces of repeated poor technical quality. In 38 of the 45 (84%) the repeat trace was satisfactory, but in the remaining seven obstetric intervention was necessary; these data are summarized in Table II. Five patients underwent induction oflabor as a result of abnormal domiciliary fetal monitoring with subsequent confirmation by cardiotocography. Two patients had their planned elective cesarean sections performed earlier. The obstetric outcome of the monitored group was similar to that of the general unit with a cesarean section rate of 16% and a forceps rate of 15%. The induction rate of28% was greater, being compared with 15%. Sixty-seven percent of the group, however, had birth weights below the 50th percentile, reflecting the recruitment of this moderately at risk grou p. The perinatal mortality rate was zero compared with 9 in 1000 for the same period in this unit. Ten patients initially referred for domiciliary fetal monitoring were eliminated from the data analysis because no trace was ever actually transmitted. Four women with postdate gestations went into labor before the first transmission. Five were incapable of operating the monitor after instruction and hence were not accepted to the monitoring program. One (a schoolteacher), having been instructed, refused to use the device because the hospital would not pay the telephone charges. Comment
Conventional unstressed fetal heart rate monitoring is well established and is the most frequently used tech-
Patient-recorded domiciliary fetal monitoring
469
Table III. Obstetric and neonatal outcome Mode of delivery Spontaneous vaginal delivery Anderson's forceps Kielland's forceps Cesarean section Elective Emergency Neonatal outcome Low Apgar score «6 at 5 min) Transferred to special care baby unit Birth weight percentiles <10th 10th-24th 25th-49th 50th-74th 75th-90th >90th Stillbirths Early neonatal deaths
89 17
4
10 13
2 3
24
26
38
32
9 5
o o
nique of fetal surveillance despite limited supportive research. The results of studies have not demonstrated universal benefit,', 6 although it has been clearly demonstrated that perinatal deaths are generally preceded by abnormal heart rate patterns. 5•S It is impossible to monitor large numbers of pregnancies with adequate frequency to detect reliably these abnormalities. The aim of domiciliary fetal monitoring recording has never been to provide any more than what is already available with conventional cardiotocography. In practical terms it is perhaps a little more difficult to interpret quantitatively the loss of baseline variability. Identification of all other abnormalities was directly comparable with that of conventional cardiotocography. Domiciliary fetal monitoring should be perceived as a screening technique that can be performed frequently but with minimal disruption to the patient and to the functioning of a perinatal unit. In this study we have looked at pregnancies with sufficient risk to require regular monitoring either directly from the clinic or after a short period of evaluation as an antenatal inpatient. High-risk pregnancies where admission would normally be required were not studied, although it may be seen in the future that these too are suitable for such surveillance. The degree of risk had been determined from the history before study entry and was confirmed to an extent by the number of babies below the 50th percentile. In spite of this risk, the overall obstetric outcome was in line with that of the unit as a whole and the fetal outcome at birth was surprisingly good with no perinatal losses. Allowing the patient to monitor her own fetal heart rate from home reduces pressure on ward, outpatient, and community midwives. One would anticipate a reduced acceptance rate. In fact, the converse is true;
470
Lindsay et al.
",hen the acceptance rates of this group of patients are compared with those of the midwife-recorded traces' in the multicenter study, those of the patients appear superior. The microprocessor technique of data storage and compression allows significant advantages over realtime techniques since telephone interference is minimized, telephone costs are reduced, and the fetal monitoring "sister" can be present at the base station for a much shorter duration of the transmission, and not the whole 30 minutes of the fetal heart recording. Although the system has been well tried in challenging situations such as distant islands off the Scottish mainland and the Harlem district of New York, our evaluation of the domiciliary fetal monitoring system has been used to assess its use in moderately "at risk" pregnancies in the Hampstead Health District. It has proved to be highly acceptable to patients, producing reassurance in almost all monitored women and indifference in a few. Acceptability and enthusiasm also have been shown by midwifery and virtually all obstetric staff. The quality of traces and the high acceptance rates appear as good in patient-recorded traces as in those performed by specifically trained midwives.' The fetal and neonatal outcome in this "at risk" group is as good as it is in the total hospital population. This has been achieved without an increase of operative obstetric intervention; the induction rate of 28%, not surprisingly, was higher. Domiciliary fetal monitoring reduces disruption to the lives of patients and their families . It enables "at risk" patients, who should be resting, to remain at home when they might otherwise require time-consuming visits to the hospital. It provides us with the ability to monitor more patients more frequently. We believe that there would be a demonstrable reduction in morbidity, mortality, and inpatient bed occupancy. These factors would release midwifery expertise and improve our overall obstetric service. Much larger numbers of patients will be needed to demonstrate this; these studies are ongoing. Further studies to assess the value of domiciliary fetal monitoring in comparison to the standard nonstress test (twice-
February 1990
Am J Obstet Gynecol
daily fetal movement charts) as sensitive means of monitoring fetal well-being are required. From the experience of this study certain recommendations can be made. It is virtually mandatory that a dedicated telephone line with a specific number is allocated to avoid the inevitable problems with hospital switchboards. Whether the system is run from an antenatal clinic (easier for referral), a labor ward (24-hour availability), or a special unit depends on local needs. The person managing the base station should be an obstetric nursemidwife available throughout each antenatal clinic. All traces should be reported early the same day to an experienced obstetrician. We thank Huntleigh Technology PLC, particularly D. Stanger and T . Justice. We also thank P. Chalk, S. M. Tuck, P. Walker, and Professor R. W. Shaw and their obstetric patients who took part in this evaluation. REFERENCES 1. Dalton KJ, Dawson AJ, Gough NAJ. Long distance telemetry of fetal heart rate from patients' own home using telephone network. Br Med J 1983;286: 1545. 2. Gough NAL, Dawson AJ, Tomkins TJ. Antenatal fetal heart rate recording and subsequent fast transmission by a distributed microprocessor-based system. Int J Biomed Comp 1986;18:61-5. 3. Dawson AJ. Middlemiss C, Jones EM, Gough NAJ. Fetal heart rate monitoring by telephone. 1. Development of an integrated system in Cardiff. Br J Obstet Gynaecol 1988;95: 10 18-1 03. 4. James D. Peralta B, Porter S, et al. Fetal heart rate monitoring by telephone. II. Clinical experience in four centres with a commercially produced system. Br J Obstet Gynaecol 1988;95: 1024-9. 5. Flynn AM. Kelly J , Mansfield H , Needham P, O C ' onoT M, Viegas O. A randomised controlled trial of non-stress antepartum cardiotography. Br J Obstet Gynaecol 1982; 89:427-33. 6. Lumley J, Lester A, Anderson I, Renou P, Wood C. A randomised trial of weekly cardiotocography in high risk obstetric patients. Br J Obstet Gynaecol 1983 ;90:1016-26. 7. Flynn AM, Kelly J. Evaluation of fetal well-being by antepartum fetal heart rate monitoring. BT Med J 1977; I : 936-9. 8. Rochard F, Schifrin BS, Goupil F, Legrand H, Blottiere J. Sureau J. Nonstressed fetal heart rate monitoring in the antepartum period . AM J OBSTET GYNECOL 1976; 126:699706.