Cost implications of introducing a telecardiology service to support fetal ultrasound screening

Abstract

A district hospital in south-east England used a telecardiology service for fetal cardiac diagnosis alongside an existing arrangement for referring pregnant women directly to perinatal cardiologists in London for detailed fetal echocardiography. Women were identified for referral according to local protocols when having a second trimester anomaly scan. For the telemedicine referrals, the sonographers video-recorded images from the anomaly scans for transmission during monthly videoconferences. The cost of the women's antenatal care was calculated from the specialist assessment until delivery, while family costs were collected in a postal survey. Over 15 months, telemedicine was used in 52 cases, while 24 women were seen in London. The London women were more likely to have had an ultrasound abnormality (29% v 10%, P = 0.047). A telemedicine assessment of 5 min duration was more costly than an examination in London (mean cost per referral of £206 v £74, P < 0.001). However, the telecardiology service was cost neutral after 14 days and for the extended period until delivery. Travel costs for London women averaged £37 compared with £5.50 for the telemedicine referrals. Telemedicine may be useful to support perinatal cardiologists in the UK whose workloads are expanding in response to improved standards in antenatal ultrasound screening.

Introduction

Almost all pregnant women in the UK are offered an anomaly scan at 18–22 weeks gestation, which is usually performed by an obstetric sonographer in the obstetric unit of a district hospital. The anomaly scan typically examines the baby's head, spine, heart, abdominal shape and content, and limbs. The UK National Screening Committee has recommended replacing the minimum imaging standard of a four-chamber view of the heart with one that includes the outflow tracts,¹ thus detecting major lesions, such as transposition of the great arteries and tetralogy of Fallot that require major surgery. Despite local training initiatives the proportion of obstetric units who report that they routinely assess outflow tracts has remained at about 60% for the last five years,^1,2 and detection rates of major abnormalities of the heart remain low overall;^3,4 the estimated rate for England and Wales in the mid 1990s was about 23%, but with marked variation between regions.⁵

In the National Health Service (NHS) the major determinants of referral to a fetal cardiologist for detailed echocardiography are the local protocols of the referring hospitals that emphasize referral of women whose pregnancies are thought to be at an increased risk for congenital heart disease (CHD). While it is accepted that the risk of CHD in these women is about 3%, the majority of babies with CHD are not born to at-risk women. Thus, this protocol-driven strategy results in a referral pattern where only a small proportion of all the babies examined by fetal cardiologists have heart defects.⁶

Diagnostic fetal echocardiography may be undertaken in paediatric cardiology departments, isolated from obstetric services, or in tertiary fetal medicine centres. Both are sited in major city hospitals and many pregnant women make lengthy journeys to attend the regional centres. In the case of a suspected abnormality only a few days elapse between the local decision to refer and the specialist appointment, so travel arrangements have to be made quickly. For all women, there are travel costs to consider as well as incidental costs, such as child-minding and possible loss of earnings for themselves or partners.

Telemedicine offers an alternative method of delivering a fetal echocardiography service. The digitized images from obstetric ultrasound machines are suitable for electronic transmission,⁷ and the structures of a fetal heart can be visualized with clarity after 18 weeks gestation. Teleconsultations can be conducted either in realtime with a sonographer examining a heart while a remote specialist jointly views the images,^8–10 or via the transmission of pre-recorded images (i.e. store-and-forward mode).

In London, the Royal Brompton Hospital (RBH) installed a telemedicine system in the 1990s. Paediatric telecardiology links were formed with hospitals in Greece and Portugal.¹¹ In 2001, the RBH set up a telecardiology network with four district general hospitals (DGHs) in south-east England for the provision of specialist advice to clinicians in obstetric and paediatric departments. The local clinicians and managers decided on the precise role for this service in their hospital. An economic analysis covering all three telemedicine applications (fetal, neonatal and paediatric) showed that the telecardiology network was cost neutral when compared with conventional referral practice once patients were followed up over six months.¹²

In one DGH, the telemedicine system was used for fetal cardiac diagnosis alongside the existing arrangement for referring women directly to London specialist centres for fetal echocardiography. A pattern of monthly ‘virtual outreach’ clinics was established during which obstetric sonographers transmitted to a specialist, ultrasound images of fetal hearts that had been recorded when women were screened for anomalies in the second trimester of their pregnancy. The present paper describes the impact of the new service on NHS costs of antenatal care received by the women and on their personal costs. It also identifies cost-related factors that could influence the wider introduction of fetal telecardiology.

Methods

The Medway Maritime Hospital in Gillingham, where the fetal telecardiology service was used, is located 55 km from central London. The fetal ultrasound department routinely referred women to perinatal cardiologists based at the RBH. The average number of women who were delivered in Gillingham was 3810 per year.

The hospital was supplied with a videoconferencing system (model 2500, Tandberg) and connected to the ISDN network at 384 kbit/s. Training in using the telemedicine equipment was provided, as well as advanced training in fetal heart scanning. At the RBH, the existing telemedicine suite contained a videoconferencing system (model 2500, Tandberg) that was used frequently by hospital staff.

Patients

Over a 15-month period in 2001–2002, demographic details were recorded of all women referred for an assessment of their baby's heart and the women's clinical events from the time of the anomaly scan until they were delivered. The personal information included any antenatal risk factors for fetal abnormalities. The factors were of two kinds: traditional CHD risk factors (a maternal or family history of CHD; maternal dependency on insulin, anti-epilepsy or lithium therapies; or current multifetal pregnancy¹³); and risk factors for Down's syndrome (notably an adjusted risk based on a quadruple serum test result or a nuchal translucency measurement ≥3.5 mm). Diagnostic information on fetal heart abnormalities was extracted from ultrasound and echocardiography reports. The project was approved by the appropriate ethics committees.

The referred women formed two groups for the analyses: a telemedicine-referrals group and a London group as a comparator. The telemedicine group consisted of women who, at the time of their anomaly scan, were identified for echocardiographic referral according to the obstetric department's protocol for high-risk women, and women whose fetal heart images during the anomaly scan were abnormal in some respect or were poorly visualized. The London group consisted of women who were referred for the same reasons in the five months before the telecardiology service entered regular use and those who travelled to London during the following 10 months, usually during the intervening weeks between the scheduled teleconferencing sessions. It also included women whose fetuses were strongly suspected of having a heart defect. A consultation with a fetal cardiologist was held according to the hospital protocols, usually within two working days.

Health service and women's costs

Detailed resource use information was collected for each woman from the anomaly scan until delivery. Unit costs in pounds sterling (£1 ≈ €1.3 ≈ US$ 2.0) for the resource items were obtained from hospital finance departments in Gillingham and specialist centres in London. The items covered ultrasound scans, antenatal and outpatient review consultations, and prenatal inpatient admissions. The unit cost for an outpatient or antenatal clinic attendance included nursing supervision and clerical staff time, consumables and equipment, overheads and capital charges. The status of the clinician (midwife or doctor) who examined each woman at each attendance was recorded separately. The mean durations of the consultations were estimated to allow the time of these clinicians to be costed and the staff cost was added to the unit cost in the analysis. The cost for an ultrasound examination incorporated the sonographer's time as well as consumables, an administrative cost, overheads and capital charges. Originally, the costs applied to the 2001–2002 financial year, but for the present paper they were inflated to 2005–2006 prices (Table 1).¹⁴ When deriving the fetal telemedicine system cost, an annual equivalent cost covering the equipment, the installation of the ISDN lines and 17.5% tax (VAT) was calculated, with an expected lifetime for the equipment of 5 years¹⁵ and an annual discount rate of 3.5%.¹⁶ The monthly invoices from the telephone company provided details of the ISDN line rental and call charges. A mean component cost per telemedicine patient was then derived (Table 2).

Table 1

Antenatal resource items with unit costs (2005–2006 prices) and frequencies of the clinical events over the period between the specialist assessment and delivery

Antenatal resource items	Unit cost (£)*	Frequency of events in total period until delivery
		Telemedicine referrals (n = 52)		London referrals (n = 24)
		Events	Patients (n)	Events	Patients (n)
Antenatal ultrasound scan (anomaly or other scans)	36.22	52 anomaly, 74 other scans	52	39 other scans	20
Antenatal or outpatient clinic attendance (review)	75.95	580 attendances	50	206 attendances	20
Clinician consulted in clinic (mean time 15 min)
Consultant	10.94
Doctor in training	6.85
Midwife	3.83
Teleconsultation clinicians (mean time 5 min)
Specialist, obstetric sonographer and co-ordinator	8.79	55 teleconsultations	52	–	–
Prenatal maternity bed-day	240.71	57 days	19	8 days	3
Termination of pregnancy	857.68	2 procedures	2	4 procedures	4
Ultrasound scan at the specialist hospital (inclusive of specialist's mean time 20 min)	64.69	7 scans	6	27 scans	24

*Costs were provided by Gillingham hospital for 2001–02 and inflated to 2005–06

Table 2

Mean costs for the fetal telemedicine system in the district hospital (2005–2006 prices)

Components of the fetal medicine system	Mean cost (£) n = 52 referred women
Telemedicine equipment	60.99
ISDN line installation, and equipment maintenance contract	9.56
Telemedicine training and service support	10.96*
ISDN line rental and call charges (mean time 5 min per woman)	76.62*
Total mean cost per woman (inclusive of value added tax)	158.14

*Costs and charges pro rated over 12 months

Information on travel arrangements and expenditure when visiting hospital was obtained in a 10-month postal survey in Gillingham of women who were referred to the specialists and women with either CHD or Down's syndrome risk factors who were not referred. Postcode data were used to calculate mileages of car journeys¹⁷ and motoring costs were applied to the mileage.¹⁸ For other travel methods, the actual costs recorded by the women were used.

Statistical analysis

After attributing unit costs to the clinical events recorded for all women (Table 1), a mean cost per woman was estimated for telemedicine referrals and London referrals for each of three time periods: the events relating to the specialist consultation; antenatal care over 14 days inclusive of the specialist consultation; and antenatal care spanning the second and third trimesters until delivery. These follow-up periods were adopted for the evaluation of all three applications of the telecardiology service in the district hospitals.¹² The short-term period covered emergency events arising from the specialist assessment (such as a termination of pregnancy following prenatal diagnosis). Fourteen days duration was selected to minimize the weighting effect on a patient's costs of any complications associated with an emergency event. The longer-term period covering the women's care until delivery had also been used in a companion paper from this project on the costs of NHS maternity care for women with high-risk and low-risk pregnancies.¹³

As the patient total costs were skewed, a simulation technique known as bootstrapping¹⁹ was used, taking 5000 iterations of the data, to calculate 95% confidence intervals around the mean costs. Statistical analyses were conducted using the software Stata version 10²⁰ and S-PLUS.²¹

Results

Over 15 months 76 Gillingham women were referred for a fetal cardiac assessment, a referral rate of 16 per 1000 maternal deliveries. Fifty-two were assessed via telemedicine and 24 travelled to London. Thirteen London women were referred before the telecardiology service entered regular use and 11 women over the following 10 months. The two cohorts were similar in terms of mean age, parity and duration of gestation (Table 3). One woman was pregnant with twins. The proportion of women with CHD-risk factors was higher in the telemedicine cohort (79%) than in the cohort referred directly to London (46%, P = 0.016). The London women were more likely to have had an ultrasound abnormality (7/24 versus 5/52, Fisher's exact test, P = 0.047). Of the 64 women referred for screening according to the departmental protocol, 47 (73%) were assessed during videoconferencing sessions. As well as following up the pregnancies until delivery, admissions to the neonatal unit were monitored and no missed diagnosis was detected among the screened babies.

Table 3

Demographic characteristics of women referred to fetal cardiology specialists over 15 months from Gillingham hospital

Characteristics of referred women	Referral method
Characteristics of referred women	Telemedicine service n = 52	Direct to London n = 24
Mean age (SD)	29 (7)	30 (7)
Parity
Primaparous	19 (37%)	12 (50%)
Multiparous	33 (64%)	12 (50%)
Risk factors for fetal abnormalities
CHD factors	41 (79%)	11 (46%)
Down's syndrome factors	5 (10%)	7 (29%)
No apparent factor	6 (12%)	6 (25%)
Timing of anomaly scan
Gestation ≤ 20 weeks	16 (31%)	5 (21%)
Gestation = 21 weeks	28 (54%)	12 (50%)
Gestation ≥ 22 weeks	8 (15%)	7 (29%)

CHD = congenital heart disease

Statistical tests: Age: t-test, P = 0.56; Parity: chi-square test, P = 0.27; Gestation week: chi-square test, P = 0.33; and CHD-risk women assessed via telemedicine: Fisher's exact test, P = 0.016

The Gillingham ultrasound department adopted the store-and-forward method for using the telemedicine service. During 10 months, 10 pre-arranged videoconferencing sessions were held. Obstetric sonographers video-recorded fetal heart images during the anomaly scans of women and the recordings were transmitted during the videoconferences. On one occasion an urgent teleconsultation was arranged, and on three other occasions, a serious heart abnormality was diagnosed during an anomaly scan performed within 48 hours of a videoconference session, so the images were transmitted for verification. About five video recordings were transmitted per session (range 2 to 8, median 5) and the mean time spent discussing a case was five minutes. Diagnostic confirmation was provided for the five cases where an abnormality had already been detected; the cases were hypoplastic left heart (3), coarctation of the aorta with mild aortic stenosis (1) and mid-muscular ventricular septal defect (1). (Four of these mothers then travelled to London for a diagnostic scan and counselling about the diagnosis, planned management and prognosis.) Five women with normal babies were re-assessed by the specialists either personally (two women) or by telemedicine (three women) to exclude progressive valvular disease because of a family history of pulmonary or aortic stenosis.

Antenatal costs of the referral services

A telemedicine assessment was significantly more costly than an examination in London (bootstrapped mean costs per referral of £206 versus £74, P < 0.001) (Table 4). This was a predictable result since the mean cost per woman of the telemedicine service in its first year of operation was £158 (Table 2). However, there was no significant difference in the antenatal mean costs for the two cohorts in each follow-up period. Six women had a late termination of pregnancy; two in the telemedicine group (4%) and four among the directly referred women (17%). These procedures were assigned a unit cost of £858, and this outcome for the pregnancies of four London-referred women contributed to the higher mean cost for the London cohort within the 14-day period. Among all the referred women, 25 were under treatment for diabetes or epilepsy; 21 in the telemedicine group (40%) and four (17%) in the directly-referred group (chi-square test, P = 0.041). A separate analysis of the study's pregnancy data-set found that the antenatal care for women with underlying medical conditions was significantly more expensive than antenatal care for singleton women in general.¹³ Thus the differing risk profiles of the telemedicine and London cohorts would account for the higher mean cost for the telemedicine women for the extended period until delivery.

Table 4

Bootstrapped mean NHS cost per patient for telemedicine and London referrals over three time periods (2005–2006 prices)

Time periods for antenatal care		Telemedicine referrals n = 52	London referrals n = 24	P value*
Day 1: Initial specialist assessment (inclusive of the anomaly scan for the telemedicine referrals)	Mean (SD) 95% CI	£206 (£12) £204 to £213	£74 (£20) £68 to £86	<0.001
14-day period (inclusive of Day 1)	Mean (SD) 95% CI	£267 (£180) £232 to £345	£348 (£516) £198 to £657	0.46
Total period from assessment on Day 1 until delivery	Mean (SD) 95% CI	£1469 (£831) £1280 to £1743	£1181 (£584) £983 to £1447	0.093

CI = confidence interval

*t-tests conducted on non-bootstrapped means

Family costs

Details of travel arrangements when attending Gillingham hospital were included in 23 questionnaires from women whose anomaly scans were video-recorded for a telemedicine assessment and 98 questionnaires from women managed locally. The mean travel costs for the two groups in 2005–2006 prices were similar: £5.49 (SD 5.03) for telemedicine referrals and £7.01 (SD 5.51) for local care. Six women were seen in London and their mean travel cost of £37.33 (SD 12.06) was significantly greater (F test, P < 0.001). Twenty-one (17%) of the Gillingham hospital attenders reported a loss of income or an expenditure on babysitting, bringing their total mean cost to £59.96 (SD 45.53).

Sensitivity analysis

A sensitivity analysis was used to examine the effect on the initial specialist assessment when personal travel costs of the referred women were added to the NHS resource costs. (Multiple imputation²² was performed to estimate the costs of women who had not been surveyed.) The bootstrapped mean cost for the telemedicine group remained significantly greater than the mean cost for the London group: £213 (SD 13) versus £108 (SD 27), t-test, P < 0.001. Sharing the telemedicine system with other users was the scenario with the greatest potential for moderating the cost of a teleconsultation.^12,15 So, as the Gillingham telecardiology service was also used in the neonatal unit, the NHS resource costs for 52 women and nine babies seen by telemedicine were compared with the costs for 24 women and eight babies who travelled to London. The telemedicine initial assessment cost was significantly less than the London assessment: £335 (SD 396) versus £1131 (SD 2708), t-test, P = 0.027. This was due to the high cost of neonatal ambulance transfers. However, at 14 days and six months, the alternative referral services were cost neutral: the six-month mean costs were £5303 (SD 14,176) for the telemedicine group and £6650 (SD 11,500) for the London group, t-test, P = 0.65).

Discussion

In the 1980s, as obstetric ultrasound techniques advanced to enable most forms of CHD to be detected in fetal life, a pattern of practice developed in the UK for hospital obstetric departments to refer women with an increased risk for fetal CHD to tertiary units for fetal echocardiography. For instance, in 1987, almost 1000 high-risk patients were referred to a tertiary centre at Guy's Hospital in London.²³ A national report on heart disease in 2002 noted that referrals for fetal echocardiography were increasing as obstetric departments' routine anomaly scans (performed at 18 to 22 weeks) had become more searching.²⁴ Perinatal or fetal cardiology consultations in England are mostly held in the 12 regional congenital cardiac units (or their associated fetal medicine centres). When the telecardiology service for this project was introduced in the four DGHs in 2001, only the Gillingham obstetric department, which was already referring high-risk women to the RBH cardiac unit, took up the service primarily for obtaining assurance of normality in ultrasound images of the fetal heart from high-risk women who would otherwise have travelled to London for echocardiography. The referral arrangements for the obstetric departments in two of the other hospitals were already established with other congenital cardiac units in London; these were unchanged and a total of 161 women were referred during our study period. In the fourth hospital, detailed fetal ultrasonography was performed within the obstetric department, so in general, relatively few women were referred (11 cases over the period).¹²

From the perspective of a district hospital, the most important telecommunications factor is the annual volume of use. The expected lifetime for the telemedicine equipment was five years,¹⁵ so the mean patient cost was based on the first year of the equipment's use plus the ISDN line rental and call charges for the same period. When the expense of the videoconferencing equipment in Gillingham was shared among the 52 women and nine babies, the adjusted mean cost (in 2005–2006 prices) for the service in the first year was reduced to £135 per patient. Senior ultrasonographers discussed pre-recorded videoed images of fetal hearts with the fetal cardiologists, spending five minutes on each case. If the women had been referred directly to the London clinics, as new patients they would have been booked for a 45-minute appointment to allow time for counselling should an abnormality be detected.

The magnitude of the financial savings to women who avoided travelling to London was determined primarily by the distances of the hospitals from their homes. The average return journey to the London clinics was 145 km compared with 21 km to the local hospital.

By assessing comprehensively the antenatal care received by the surveyed women during the second and third trimesters of their pregnancy, we observed how a teleconsultation was a relatively modest cost component of the total package of care provided by the NHS. Moreover, the telemedicine referral service was cost neutral in the extended period until delivery. Analyses of the study's data-sets showed that the total antenatal mean cost for a cohort of women could be markedly affected by costs associated with their pre-existing or pregnancy-related medical conditions.¹³ Thus, in the telemedicine group, almost 80% of the 52 women had CHD risk factors, such as diabetes, epilepsy or a family history of CHD, and 19 (37%) experienced a prenatal admission to hospital at a cost of £241 per bed-day.

Since the study, the fetal telecardiology service in Gillingham has continued to use the same videoconferencing system. During 2006, eight telemedicine sessions of 45–60 min duration were held and the videos of 84 women were assessed. Eleven of these women were followed up by London specialists and the hearts of four babies proved to be abnormal. Maternal obesity can impede visualization of the fetal heart during an anomaly scan. More than half the fetal cardiac images transmitted were from women with a body mass index above 30 but were of sufficient quality to distinguish abnormal cases. The ultrasound department remained committed to the telemedicine service because of its educational value for training staff in addition to its clinical effectiveness. Counselling is currently undertaken face to face in London as it is felt that this is better for achieving mutual understanding between the specialist and parents when major decisions, such as a termination of a pregnancy, are made, and specialist fetal cardiac liaison nurses are available to provide parental support. (It is also relevant that fetal images from specialist echocardiography machines are of a higher quality than images from obstetric ultrasound machines.) In the longer term, however, it may become acceptable to counsel parents during teleconferences, especially for more straightforward CHD lesions.

Only a small proportion of the 75 paediatric cardiology consultants in England are also specialists in fetal cardiology, so if their workloads are to be kept in check, a change of practice is needed to reduce the amount of time spent by them on the great majority of high-risk women whose fetuses are normal. The number of women with fetal cardiac risk factors in the UK is likely to increase as more and more survivors of CHD (both males as potential fathers, and females) reach adulthood, and the incidence of type 2 diabetes continues to rise.²⁵ As maternal obesity becomes more prevalent, a subpopulation with poor fetal heart imaging is likely to contribute to the workload.

New challenges to the existing workload of fetal cardiologists will arise as standards of antenatal screening in maternity services rise,²⁶ particularly as visualization of outflow tracts of the fetal heart is increasingly performed during second trimester anomaly scans.¹ Higher detection rates of equivocal and unequivocal cardiac anomalies by obstetric sonographers will, in turn, lead to appropriately higher rates of referral for perinatal echocardiography. Nuchal translucency screening in the first trimester is already available in one-quarter of English hospital trusts² in accordance with recently issued national guidelines.²⁷ This will substantially increase the number of women requiring an early and more complex fetal cardiac scan performed by a specialist, given the independent higher risk of CHD despite a normal karyotype.

In conclusion, the present study demonstrates the potential for fetal telecardiology to support the few perinatal cardiologists in the UK in providing timely specialist consultations for women with frank or suspected fetal abnormalities. Setting up networks between specialist units and referring district hospitals for the transmission of pre-recorded video or digitally-stored ultrasound images, where up to 12 patients can be discussed in hour-long videoconference sessions, is likely to be an efficient and economic way of providing support and training to local obstetric screening programmes and will avoid unnecessary referral to tertiary centres.

Footnotes

Acknowledgements

We are grateful to staff at the obstetric department of Medway Maritime Hospital in Gillingham for participating in this study. We thank Helen Kennett, Daphne Brown, Dr Mike Rigby and Dr Gwyn Weatherburn for assistance. Funding for the study was from the Policy Research Programme of the Department of Health and the Charitable Funds Committee of the Royal Brompton and Harefield NHS Trust. The views expressed in the publication are those of the authors and not necessarily those of the funders.

References

UK National Screening Committee. Antenatal Ultrasound Screening. Ultrasound Survey of England: 2002. See: http://www.screening.nhs.uk/fetalanomaly/ultrasound_survey.pdf (last checked 6 September 2008)

Healthcare Commission. Key findings of the 2007 Maternity Service Review. See: http://www.healthcarecommission.org.uk/_db/_documents/Maternity_FINAL_BRIEFING_NOTE.doc (last checked 6 September 2008)

Bricker

, Garcia

, Henderson

, Ultrasound screening in pregnancy: a systematic review of the clinical effectiveness, cost-effectiveness and women's views. Health Technol Assess 2000;4:1–193

Acherman

, Evans

, Luna

, Prenatal detection of congenital heart disease in southern Nevada: the need for universal fetal cardiac evaluation. J Ultrasound Med 2007;12:1715–19

Bull

. Current and potential impact of fetal diagnosis on prevalence and spectrum of serious congenital heart disease at term in the UK. Lancet 1999;354:1242–7

Kilby

. Fetal Medicine. See http://www.bwhct.nhs.uk/print/index/about_clinicalreport_intro/about_clinical_report-200304_intro/about-clinicalreport_200304/about-clinicalreport-fetalmed.htm (last checked 25 September 2008)

Begg

, Chan

, Edie

, Hockey

, Wootton

. Minimum acceptable standards for digital compression of a fetal ultrasound video-clip. J Telemed Telecare 2001;7 (Suppl. 2):88–90

Fisk

, Sepulveda

, Drysdale

, Fetal telemedicine: six month pilot of real-time ultrasound and video consultation between the Isle of Wight and London. Br J Obstet Gynaecol 1996;103:1092–5

Chan

, Soong

, Watson

, Whitehall

. Realtime fetal ultrasound by telemedicine in Queensland. A successful venture? J Telemed Telecare 2001;7 (Suppl. 2):7–11

10.

Sharma

, Parness

, Kamenir

, Screening fetal echocardiography by telemedicine: efficacy and community acceptance. J Am Soc Echocardiogr 2003;16:202–8

11.

Tsilimigaki

, Maraka

, Tsekoura

, Eighteen months' experience with remote diagnosis, management and education in congenital heart disease. J Telemed Telecare 2001;7:239–43

12.

Dowie

, Mistry

, Young

, Telemedicine in pediatric and perinatal cardiology: economic evaluation of a service in English hospitals. Int J Technol Assess Health Care 2007;23:116–25

13.

Mistry

, Dowie

, Young

, Costs of NHS maternity care for women with multiple pregnancy compared with high-risk and low-risk singleton pregnancy. BJOG 2007;114:1104–12

14.

Curtis

, Netten

. Pay and Prices Index. Unit Costs of Health and Social Care 2006. Canterbury: Personal Social Services Research Unit, University of Kent at Canterbury, 2006

15.

Jacklin

, Roberts

, Wallace

, Virtual outreach: economic evaluation of joint teleconsultations for patients referred by their general practitioner for a specialist opinion. BMJ 2003;327:84–8

16.

HM Treasury. The Green Book: Appraisal and evaluation in central government. See http://www.hm-treasury.gov.uk/media/9/C/Green_Book_03.pdf (last checked 6 September 2008)

17.

Multimap. See http://www.multimap.com/ (last checked 6 September 2008)

18.

Automobile Association. Archive of Running Costs 1998–2007. See http://www.theaa.com/motoring_advice/running_costs/archive.html (last checked 6 September 2008)

19.

Manly

BFJ

. Randomisation, Bootstrap and Monte Carlo Methods in Biology. 2nd edn. London: Chapman and Hall, 1997

20.

StataCorp. Stata Statistical Software: Release 10.0. College Station, TX: Stata Corporation, 2007

21.

MathSoft, Inc. S-PLUS 2000 Professional Release 2. Seattle, WA: MathSoft, Inc., 1999

22.

Rubin

. Multiple Imputation for Non-response in Surveys. New York, NY: Wiley, 1987

23.

Allan

, Chita

, Sharland

, Fagg

, Anderson

, Crawford

. The accuracy of fetal echocardiography in the diagnosis of congenital heart disease. Int J Cardiol 1989;25:279–88

24.

Anon. Fifth report on the provision of services for patients with heart disease. Heart 2002;88 (Suppl. 3):1–56

25.

Feig

, Palda

. Type 2 diabetes in pregnancy: a growing concern. Lancet 2002;359:1690–2

26.

Harcombe

, Fairgrieve

. National training needs analysis for antenatal screening services across England. See http://www.perinatal.nhs.uk/screening/resources/Training_%20Needs_Analysis.pdf (last checked 6 September 2008)

27.

UK National Screening Committee. Fetal Anomaly Screening Programme. See http://www.nsc.nhs.uk/hottopics/hottopics_ind.htm (last checked 6 September 2008)