5.1 The sponsor submitted a de novo cost analysis evaluating the cost consequences of using the Vision Amniotic Leak Detector (ALD) compared with speculum examination alone. The population was pregnant women with unexplained vaginal wetness that could be leaking amniotic fluid caused by ruptured membranes. The model was a simple decision tree comparing 2 pathways:
use of speculum examination in the entire population presenting to an antenatal day unit with unexplained wetness
use of the Vision ALD to provide an initial diagnosis, followed by speculum examination in women with a positive Vision ALD result and discharge for women with a negative result.
The sponsor did not specify a time horizon because consequences were not considered beyond initial examination and diagnosis in the day unit. The sponsor included base-case, worst-case and best-case scenarios in their sensitivity analysis.
5.2 The key assumptions used in the model were:
The Vision ALD had 100% sensitivity.
Speculum examination had 100% sensitivity and specificity.
All women receiving speculum examination would need cardiotocography (CTG; fetal heart rate monitoring), but women in the Vision ALD arm with a negative result would not receive CTG.
Women needing speculum examination would wait on a bed.
Women would not wait on a bed before using the Vision ALD.
The Vision ALD would be used as the sole diagnostic test to rule out ruptured membranes.
5.3 Clinical parameters from the study by Mulhair et al. (2009) were used to inform the model, including the reported percentage of women with a negative Vision ALD result (38%, n=53/139); however, this was wrongly transcribed as 42%, which was the reported prevalence. The model did not account for false positive Vision ALD results; the study by Mulhair et al. (2009) identified 12 false positives (9%) that could not be attributed to infection or subsequent delivery.
5.4 The cost of the Vision ALD was £1.60 and the cost of a disposable speculum was £0.84 (provided by the sponsor). The cost of bed use in the antenatal unit was estimated to be £364, taken from the healthcare resource group for antenatal false labour including prelabour rupture of membranes (PROM) (NZ23Z; excess bed day cost of £364). The total cost of bed use in the antenatal day unit for speculum examination (a total of 35 minutes) was £8.90, based on 30 minutes pooling time plus the examination. The costs of infection identification and treatment were not included in the model.
5.5 The cost per minute of midwife time was estimated as £1.37, extrapolated from £82.00 per hour (Personal Social Services Research Unit [PSSRU] 2011). The sponsor estimated the times for administering the Vision ALD and a speculum examination to be 5 minutes each, with a total cost of £6.85 for midwife time for each test. The total cost per use of the Vision ALD was therefore estimated to be £8.45 and for speculum examination, the total cost per use was estimated to be £43.99. The cost of CTG was added to the cost of the speculum examination, amounting to an estimated 20 minutes of midwife time (£27.40).
5.6 The sponsor carried out one-way and multi-way deterministic sensitivity analyses to produce base-case, worst-case and best-case cost scenarios. The costs of the Vision ALD and a disposable speculum were fixed. Bed day cost was varied according to a published range. All other variables were adjusted to plus or minus 50% of the base case.
5.7 The sponsor carried out a threshold analysis to determine at what point the Vision ALD became cost neutral, by varying midwife time to conduct a speculum examination and CTG, and midwife time to administer the Vision ALD, as well as the cost of the Vision ALD.
5.8 The sponsor's base-case analysis showed that the cost per woman of using the Vision ALD, followed by speculum examination in the event of a positive result, was £33.93. The cost per woman of using speculum examination alone was £43.94, thus generating a cost saving per woman of £10.01. The results of the one-way sensitivity analysis carried out by the sponsor showed that the Vision ALD remained cost saving for each of the varied parameters. Cost savings ranged from £0.78 to £19.23, the broadest variation being related to the percentage of negative Vision ALD tests (63% to 21%).
5.9 The sponsor's best- and worst-case scenarios for the Vision ALD generated, respectively, a cost saving of £54.60 and an additional cost of £3.52 per woman. The threshold analysis showed that the Vision ALD would no longer be cost saving if the time taken to perform a speculum examination and CTG was reduced to 7.6 minutes or less, the time taken to administer the Vision ALD was increased to 12.3 minutes, the cost of the Vision ALD was increased to £11.60 or the percentage of women with a negative result was reduced to 19%.
5.10 The External Assessment Centre considered that the sponsor's de novo analysis was limited in that it did not include the cost consequences of using the Vision ALD in a community setting and it did not explore the full diagnostic pathway. The External Assessment Centre carried out additional modelling to address these limitations.
5.11 The External Assessment Centre's model consisted of a decision tree with 2 main arms, comparing standard care for diagnosing unexplained vaginal wetness (speculum examination), with use of the Vision ALD, followed by speculum examination in the event of a positive result. In the Vision ALD arm, women with a negative test result would be sent home with no further intervention. Each arm of the model was populated separately for PROM or preterm prelabour rupture of membranes (PPROM) to account for the differences in management and the prevalence of infection in these 2 groups.
5.12 The standard care arm included 3 usage scenarios based on a cohort of 1000 women, in which speculum examination was carried out by a GP, a midwife based in a GP practice or a clinician in an antenatal day unit. In the Vision ALD arm, women were given the Vision ALD test in the same clinical settings and, in addition, by a community midwife in the woman's home.
5.13 The health outcomes considered in the model were:
diagnosis of ruptured membranes
diagnosis of ruptured membranes with infection
diagnosis of infection without ruptured membranes
no infection or ruptured membranes.
5.14 The proportion of positive and negative speculum and Vision ALD tests were taken from the External Assessment Centre's additional analysis of the published studies by Bornstein et al. (2006 and 2009) and Mulhair et al. (2009), described in sections 3.2–3.5. A weighted mean rate of positive tests was calculated. The weighted mean positive predictive value of the Vision ALD was also used to determine the number of positive PROM or infection diagnoses after a speculum examination in the Vision ALD arm.
5.15 The infection rates for PROM (1%) and for pregnant women in general (0.5%) were taken from the NICE clinical guideline on intrapartum care. For PPROM, a rate of 28% was taken from the Royal College of Obstetricians and Gynaecologists' guidance (2006).
5.16 Key assumptions used in the model were:
Antibiotic therapy would be given for 24 hours, based on the assumption that delivery would occur within 24 hours in most women with PROM. The External Assessment Centre noted that for PPROM, delivery could occur after 24 hours, but opted to use the same antibiotic costs for both conditions to simplify the model.
Each usage scenario would follow a like-for-like pattern with care being provided in the same setting throughout the pathway. The only exception to this would be in the community midwife scenario for the Vision ALD, when the woman would be referred to a midwife in a GP practice for the speculum examination.
The clinical times for administering the Vision ALD (10 minutes) and carrying out a speculum examination (15 minutes) included time for consultation and discussion.
False negatives were not included in the model, because the rate of false negative results calculated by the External Assessment Centre (around 4%) was similar for both tests.
The cost of CTG was not included in the model because this would only be available in an antenatal day unit and would most likely be accounted for in a referral cost.
5.17 Costs were estimated for clinical time and resource use, as well the treatment and monitoring of infection and PROM or PPROM. The costs for the Vision ALD (£1.60) and a disposable speculum (£0.84) were taken from the sponsor's model. Clinical time costs were taken from PSSRU 2012 unit costs, giving a per minute rate of £3.68 for a GP, £0.88 for a midwife based in a GP practice and £1.18 for a community midwife. No costs were applied for resource use (for example, a bed or consultation room) for speculum examination in the GP practice. The External Assessment Centre stated that the cost for this was most likely to be negligible, based on calculations using GP cost elements (PSSRU 2012, section 10.8a) and would not significantly affect the outcomes in the model. The External Assessment Centre also stated that availability of unused space would vary considerably between practices and so generalisation would be difficult to include in the model.
5.18 The costs of treating infection consisted of microbiology laboratory costs of £8.00, taken from NHS reference costs, and broad spectrum antibiotic costs of £41.72 for 24 hours, based on published evidence. A small cost was applied for temperature monitoring to identify potential infection after a diagnosis of PROM or PPROM, based on NHS supply chain costs for a Tempa-dot thermometer at £0.90 for 10. The External Assessment Centre's sensitivity analysis also included costs for a 24-hour course of antibiotics specifically for group B streptococcus at a cost of £22.71; this was added on the basis of clinical advice. Costs for managing PPROM consisted of corticosteroids at a cost of £4.98 and a course of prophylactic antibiotics (if no infection was diagnosed) at £2.41. A cost of £696.00 was applied for inpatient observation after a diagnosis of infection and/or PPROM. The cost for referral to an antenatal day unit was £81.00; both figures were taken from NHS reference costs. The External Assessment Centre assumed that the cost of an antenatal day unit referral would cover clinical time and resource use.
5.19 The External Assessment Centre found that use of the Vision ALD was cost-incurring compared with standard care, for both PROM and PPROM, in all the clinical usage scenarios described in section 5.12. The cost of the Vision ALD device and the additional clinical time taken to administer it were not outweighed by the savings from avoided speculum examinations across the 1000-woman cohort. Administration of the Vision ALD by a GP would incur the greatest cost (£14.85 per woman) and the least cost-incurring scenario would be its use in the antenatal day unit (£1.28 per woman).
5.20 The External Assessment Centre carried out a one-way sensitivity analysis, altering each clinical time and treatment cost. Costs were estimated for 'lowest' and 'highest' ranges. The 2 variables with the greatest impact on both arms of the model were clinician time taken to administer the Vision ALD and to carry out speculum examination (by a GP or midwife). Threshold analysis on these 2 variables was carried out to determine at what level the Vision ALD would become cost saving. For PROM, the time taken to carry out speculum examination would need to be around 30 minutes and the time taken to administer the Vision ALD would need to be 6 minutes or less, both varying depending on the clinician carrying out the test. Similar figures were found for PPROM.
5.21 The External Assessment Centre then explored the cost consequences, using their existing model, of additional scenarios for the Vision ALD in which women attended 2 different facilities. The scenarios modelled were a GP or midwife issuing the Vision ALD in a GP practice, or a community midwife issuing the Vision ALD in the woman's home, followed by referral to either a practice-based midwife or GP, or an antenatal day unit, for a speculum examination. For the standard care arm, all women would have a speculum examination by a practice-based midwife or GP, or at an antenatal day unit. The setting for the speculum examination would be the same in both arms for each scenario compared.
5.22 Base-case results indicated that cost savings could be achieved if the clinician issuing the Vision ALD is of lower cost than the clinician carrying out the speculum examination, because the cost saving was driven by avoided speculum examinations in high-cost settings. A GP issuing the Vision ALD, followed by referral to a practice-based midwife or antenatal day unit for a speculum examination, incurred costs of between £3.99 per woman in the PROM arm and £38.28 per woman in the PPROM arm. A practice-based midwife issuing the Vision ALD and referring women to an antenatal day unit for speculum examination was associated with estimated cost savings of £24.01 per woman for PROM and £18.25 per woman for PPROM. A practice-based midwife issuing the Vision ALD followed, if required, by speculum examination by a GP, could generate estimated cost savings of £13.15 per woman for PROM and £7.40 per woman for PPROM. The scenario of a community midwife issuing the Vision ALD in the home and referring women to an antenatal day unit for speculum examination could generate cost savings of £21.01 and £15.25 per woman for PROM and PPROM respectively; referral to a GP could generate savings of £10.15 per woman for PROM and £4.40 for PPROM.
5.23 The Committee considered the scenarios and findings from the sponsor's model and the additional analyses carried out by the External Assessment Centre. The Committee heard advice from clinical experts regarding which scenarios were likely to reflect a realistic care pathway. The advice indicated that the most likely scenarios in the community were use of the Vision ALD by a practice-based or community midwife, followed by referral to an antenatal day unit in the event of a positive Vision ALD result. The Committee noted that in these scenarios, use of the Vision ALD was cost saving in the External Assessment Centre's analysis.
5.24 The Committee noted that the observed cost savings were largely a result of avoiding speculum examinations in women with a negative Vision ALD test. It also noted that a significant factor in the analysis was the cost of the healthcare provider administering the test.
5.25 The Committee noted that use of the Vision ALD in an antenatal day unit was a likely scenario based on current practice, and that this had been shown to be cost-incurring in the External Assessment Centre's model. The Committee considered that use in this setting would become cost neutral if the consumable costs of the Vision ALD and disposable speculums were included in the reference cost of a referral.