- Top of page
Objectives To assess the cost-effectiveness of outpatient (at home) cervical ripening with isosorbide mononitrate (IMN) prior to induction of labour.
Design Economic evaluation was conducted alongside a randomised placebo controlled trial (the IMOP trial).
Setting Large UK maternity hospital.
Population A total of 350 nulliparous women with a singleton pregnancy, cephalic presentation ≥37 weeks gestation, requiring cervical ripening prior to induction of labour.
Interventions Isosorbide mononitrate (n = 177) or placebo (n = 173) self-administered vaginally at home at 48, 32 and 16 hours prior to the scheduled time of admission for induction.
Results Mean health service costs between the period of randomisation and discharge for mother and infant were £1254.86 in the IMN group and £1242.88 in the placebo group, generating a mean cost difference of £11.98 (bootstrap mean cost difference £12.86; 95%CI: −£106.79, £129.39) that was not statistically significant (P = 0.842). The incremental cost per hour prevented from hospital admission to delivery was £7.53. At the notional willingness to pay threshold of £100 per hour prevented from hospital admission to delivery, the probability that IMN is cost-effective was estimated at 0.67. This translated into a mean net monetary benefit of £98.13 for each woman given IMN.
Conclusions Although the probability that IMN is cost-effective approaches 0.7 at seemingly low willingness to pay thresholds for an hour prevented from hospital admission to delivery, our results should be viewed in the light of the clinical findings from the IMOP trial.
- Top of page
Post-term pregnancy is associated with increased rates of fetal complications, including meconium and meconium aspiration syndrome, oligohydramnios, macrosomia, fetal birth injury, non-reassuring fetal heart rate or fetal distress in labour and caesarean delivery.1 Labour induction, carried out to reduce the risk of complications in post-term pregnancies, is successful depending on the degree of cervical ripening.2 Isosorbide mononitrate (IMN) is a nitric oxide (NO) donor that can be used for cervical ripening without causing uterine contractions and the need for fetal monitoring, and therefore be safely administered at home.2,3 Randomised controlled trials (RCTs) have been published on the use of prostaglandins (PG), another cervical ripening agent, but the agent has since been declared unsafe for the outpatient (at home) context by many authorities because of adverse fetal and maternal effects.4,5 There is only one randomised controlled study, conducted in Sweden and published in early 2007, on the use of IMN for cervical ripening on an outpatient basis.6 That study found that outpatient administered IMN is clinically effective, safe and well tolerated, although it provided no evidence of the cost-effectiveness of the agent. In this paper, we present the results of an economic evaluation of outpatient (at home) cervical ripening using IMN based on the Isosorbide MOnonitrate against Placebo study (IMOP) trial. This economic evaluation has been written in conjunction with the paper reporting clinical outcomes from the IMOP trial.7
- Top of page
A total of 350 women were randomised in the trial; 177 to the active group and 173 to the placebo group. The groups were similar in terms of mean maternal age, mean gestational age at delivery, number of previous pregnancies, indication for labour induction and modified bishop score.7
Isosorbide mononitrate reduced the hospital admission to delivery interval as compared with placebo by an average of 1.6 hours (95% CI: −1.9, 5.1; P = 0.37) (Table 1). There were no significant differences in mode of delivery, length of labour, oxytocin requirements during labour and the use of epidural analgesia in labour between the trial groups. The proportion of women with an unripe cervix (defined as Bishop score of <7) after 24 hours of outpatient treatment was significantly lower in the IMN group as compared with the placebo group (64% versus 77%; P = 0.021). The proportion of women requiring further PG was also lower in the IMN group as compared with placebo group (68% versus 77%), but this did not reach statistical significance (P = 0.065). There was no significant difference in neonatal outcomes with respect to the proportion of babies admitted to the neonatal unit.
Table 1. Primary and secondary outcomes for mothers and babies randomised in IMOP trial
| ||IMN (n = 177)||Placebo (n = 173)|
|Elapsed time interval from hospitaladmission to delivery (hours)|| 25.06|| 26.66|
|Secondary maternal and neonatal outcomes (%)|
|Lower segment caesarean sections (LSCS)|| 65 (37)|| 56 (33)|
|Length of labour (hours)|| 8.38|| 8.14|
|Oxytocin augmentation rates||127 (72)||127 (73)|
|Epidural usage||116 (65)||120 (69)|
|Number with unfavourable cervix at 24 h after admission|| 83 (64)|| 98 (77)|
|Number with requirement for additional inpatient cervical ripening agent|| 87 (68)|| 98 (77)|
|Frequency of neonatal admissions to special care|| 18 (10)|| 16 (9)|
|Duration of neonatal admission (hours)|| 2.57|| 0.47|
|Still birth or neonatal death|| 0|| 1|
Health service resource use
Women allocated to the IMN group spent, on average, a fewer number of hours on the prenatal assessment ward (where they are admitted when they arrive into hospital before labour has commenced) (P = 0.32) and in the labour ward (P = 0.63) than women allocated to the placebo group (Table 2), but none of these differences were statistically significant. Likewise, length of stay on the postnatal ward was not significantly different between the randomised groups (P = 0.40), nor were there any significant differences in the rates of maternal complications or neonatal hospitalisation (including ambulance transfers and length of stay in intensive care and high dependency units), apart from average length of stay in the special care baby unit (2.57 hours in the IMN group versus 0.47 hours in the placebo group; P = 0.05). The reasons for neonatal hospitalisation in special care (18 and 16 in the IMN and placebo groups respectively) included meconium aspiration, Group B streptococcal infection, respiratory distress, and neonatal pneumonia and sepsis, none of which are considered to have biological plausibility for attribution to IMN. The cost of neonatal complications experienced by infants was reflected in the cost calculations underpinning the Department of Health reference costs.10
Table 2. Resource use and unit costs of resource items
|Resource use variable||Resource use value||Unit cost or range**|
|IMN (n = 177)||Placebo (n = 173)|
|Length of stay in pre natal ward (min)||998.29 (986.73)||1100.79 (985.96)|| 0.33 per mina|
|Length of stay in labour ward (min)||522.48 (232.35)||541.59 (221.39)|| 0.75 per mina|
|Length of stay in postnatal ward (day)||2.71 (1.37)||2.58 (1.23)||150.63 per daya|
|Length of stay in intensive care ward (day)||0.03 (0.21)||0.05 (0.28)||711.85 per dayb|
|LSCS||112.42 (121.64)||103.97 (117.88)||215.18a|
|Assisted deliveries*|| 47 (27%)|| 54 (31%)|| 44.01–215.18 per deliverya, c|
|PPH, tears and wound and urine infections||0.34 (0.47)||0.29 (0.45)|| 7.83–142.48 per complicationa,d|
|Ambulance transfer (episodes)||0.01 (0.76)||0.01 (0.76)||298.68 per transferb,e|
|Length of stay in intensive care unit (day)|| (0.01) (0.75)|| (0.01) (0.00)||972.57 per dayb|
|Length of stay in high dependency unit (day)||0.01 (0.11)||0.00 (0.00)||633.11 per dayb|
|Length of stay in special care unit (day)||2.57 (13.7)||0.47 (2.83)||388.42 per dayb|
Health service costs
There were no statistically significant differences between the two groups for any cost category or total costs (Table 3). Mean health service costs between the period of randomisation and discharge for mother and infant were £1254.86 in the IMN group and £1242.88 in the placebo group, generating a mean cost difference of £11.98 (bootstrap mean cost difference £12.86; 95% CI: −£106.79, £129.39) that was not statistically significant (P = 0.842). Overall cost differences between the two groups can be largely explained by the additional care received by infants in the IMN group during the neonatal period [mean cost difference of £35.23 (bootstrap mean cost difference £33.70; 95%CI: −£2.17, £75.93; P = 0.088)] and the increased cost of postnatal care, which was higher by an average of £8.82 (bootstrap mean cost difference £8.82; 95%CI: £50.98, £64.74; P = 0.766). The largest cost saving was attributable to reduced stay in the pre natal assessment ward [mean cost difference of £33.82 (bootstrap mean cost difference −£35.38; 95%CI: −£96.89, £32.97; P = 0.332)]. The cost of labour and associated complications was a major cost component in the trial, although there was no significant difference between the groups (P = 0.977).
Table 3. Mean costs and mean differences by cost category
|Cost category||IMN (n = 177)||Placebo (n = 173)||Mean difference||P value*||Bootstrap mean difference (95% CI)**|
|Medications including IMN||12.64||26.89||10.27||23.76||2.37||0.383||2.53 (−2.93, 8.16)|
|Ambulance transfer||1.69||22.45||1.73||22.71||−.040||0.987||−0.23 (−5.18, 5.06)|
|Prenatal assessment ward||329.44||325.62||363.26||325.37||−33.82||0.332||−35.38 (−96.89, 32.97)|
|Labour and complications||443.08||186.39||443.65||188.21||−0.57||0.977||−0.87 (−40.71, 39.16)|
|Postnatal care||430.07||279.17||421.25||275.56||8.82||0.766||8.94 (−50.98, 64.74)|
|Neonatal care||37.96||272.47||2.73||13.19||35.23||0.088||33.70 (−2.17, 75.93)|
|Total NHS costs||1254.86||625.26||1242.88||487.56||11.98||0.842||12.86 (−106.79, 129.39)|
The incremental cost per hour prevented from hospital admission to delivery was estimated at £7.53. Variability around the estimate of cost-effectiveness for the principal outcome is shown in the cost-effectiveness plane displayed in Figure 1, providing a graphical presentation of the distribution of the bootstrapped cost-effectiveness pairs for IMN. Although the majority of the bootstrapped samples fall in the north east quadrant of the cost-effectiveness plane, some bootstrapped samples fall in all four quadrants, resulting in a problem when interpreting negative incremental cost effectiveness ratios (ICERs). A negative ICER might represent improved outcomes and lower costs as a result of IMN, or worse outcomes and higher costs. This means that a meaningful ordering of the bootstrapped samples, which is required to make the CI surrounding the ICER interpretable, is very difficult. Under these circumstances, cost-effectiveness acceptability curves, which summarise the uncertainty in estimates of cost-effectiveness, represent the appropriate approach to representing the uncertainty surrounding the ICER. The cost-effectiveness acceptability curve for the principal outcome measure is shown in Figure 2. This indicates the probabilities that IMN is cost-effective according to notional willingness to pay thresholds held by decision makers for an hour prevented from hospital admission to delivery. At the notional willingness to pay threshold of £100 for each inpatient hour averted, the probability that IMN is cost-effective was estimated at 0.67. This increases to 0.77 at a notional willingness to pay threshold of £1000.
The estimates of the costs and consequences of IMN reported in this paper were derived from a rigorously conducted RCT, with the uncertainty surrounding the ceiling cost-effectiveness ratio that decision-makers would consider acceptable dealt with through the use of cost-effectiveness acceptability curves. However, we also performed sensitivity analyses to determine the impact that uncertainty surrounding individual parameter values might have on the incremental cost-effectiveness for the principal outcome. Varying hospital occupancy rates to 10% greater or 10% less than revealed by March 2005 to December 2006 hospital data varied the ICER to £10.69 and £4.89 respectively. Changing the per diem costs for each level of hospital maternity care to the mean Department of Health reference costs marginally increased the incremental cost-effectiveness ratio to £9.46; varying these costs to the upper and lower inter-quartile limits reduces (increases) the incremental cost-effectiveness ratio to £0.44 (and £12.68) respectively. The sensitivity analysis for the per diem costs of each level of neonatal care, varied to the upper and lower inter-quartile limits for Department of Health reference costs, had minimal effects on the incremental cost-effectiveness ratio (£5.31 and £9.74 respectively).
Mean net benefits
Mean net benefits were estimated for alternative willingness to pay thresholds for an hour prevented from hospital admission to delivery (Table 4). Assuming that Rc equals £100 for an hour prevented from hospital admission to delivery generates a mean net benefit to the health services attributable to IMN of £98.13 (i.e., there is a net gain to the health services in monetary terms). If Rc is increased as high as £5000 for an hour prevented from hospital admission to delivery, the mean net benefit increases to £7805.
Table 4. Mean net benefit of IMN, calculated for alternative willingness to pay thresholds per unit of outcome
|Willingness to pay threshold||Hour of hospital admission averted|
|Mean net benefit||95% CI*|
- Top of page
This study demonstrates that IMN increases overall health service costs. When combined with clinical effectiveness data from the IMOP trial, IMN generates an incremental cost per hour prevented from hospital admission to delivery of £7.53, yielding a positive net monetary benefit for each additional use of IMN at willingness to pay thresholds as low as £40. This conclusion remains relatively robust following extensive sensitivity analyses that account for uncertainty surrounding the values of parameters incorporated into the economic evaluation and for uncertainty surrounding the willingness to pay threshold for the primary clinical outcome.
The probability that IMN is cost-effective approaches 0.7 at seemingly low willingness to pay thresholds for an hour prevented from hospital admission to delivery. However, our results should be viewed in the light of the clinical findings from the IMOP trial, showing that the reduction in hospital admission to delivery interval was not significant.
There are several strengths to our study. It was based on a large RCT in which bias was eliminated through random allocation of women to the study. The study presented in this paper was based on rigorous methods, which included an extensive sensitivity analysis, methods for handling uncertainty and the latest health economic evaluation methods. Moreover, the induction of labour protocol which was developed by the hospital and followed by the trial14 was generalisable to other UK settings. There are three possible limitations of this work that should be borne in mind by readers. First, although our primary outcome measure has clinical relevance when evaluating the benefits of cervical ripening agents, decision makers such as the National Institute for Health and Clinical Excellence (NICE) usually recommend preference based measures, such as quality-adjusted life years (QALYs), for economic evaluation purposes.15,16 It was considered that the QALY metric would not be relevant in this clinical setting as the potential effects of IMN are likely to be limited to a restricted time frame for both mother and infant. Indeed, it is likely that important characteristics of maternity care, such as the ability to make decisions during labour, cannot be detected by traditional health benefit measures such as QALYs.17 However, since previous research has indicated that women are able to express clear preferences for maternity care,18 willingness to pay and stated preference discrete choice experiment methods could, in principle, be used to estimate their preferences for alternative attributes of care that might be important in this setting. These preferences, in turn, could be used to inform decision makers willingness to pay values for outcomes of interventions such as cervical ripening with IMN, as well as other secondary outcomes in the trial.
A second limitation of our study was that we primarily derived our costs from the Princess Royal Maternity hospital in Glasgow. Although our cost accounting was comprehensive and included all significant cost items, whose values were calculated according to established principles in economic theory, alternative accounting methods applied in different jurisdictions may generate different cost estimates. We addressed this issue, in part, by performing sensitivity analyses that applied national reference costs to the main estimates of resource utilisation and this did not significantly affect the final cost-effectiveness results. Nevertheless, cost estimates derived using rigorous accounting methods from a broader profile of hospital settings are likely to increase the generalisability of our study results.
A final limitation was that wider costs, such as those borne to women, partners or carers, were not included in this study. Widespread implementation of IMN on an outpatient (at home) basis may increase broader societal costs, such as increased care by partners, more time off work and increased child care requirements for dependents. Alternatively, it may mean that more women are satisfied with the process of their labour and are faced with fewer personal costs. This remains to be elucidated by economic studies in this area that adopt a broader societal, rather than health service, perspective.
Economic evidence in the area of induction of women generally suggests that an unfavourable cervix prior to induction of labour significantly increases the rate of caesarean delivery.19,20 As caesarean delivery is the most costly type of delivery,21 this has important implications for the NHS. Our study found no statistically significant differences in the rate of caesarean delivery between the two groups, suggesting that IMN did not achieve a positive effect in reduced caesarean section rates.
Although we were not able to demonstrate overall cost savings attributable to IMN, we did find that women receiving IMN spent less time in the prenatal assessment ward, and this accounted for the largest cost difference between the trial groups. Similar to our findings, another study of women receiving intensive dose vaginal dinoprostone in pre-induction cervical priming has shown shortening of the period of pre-labour hospital stay for women22 although no cost-effectiveness analysis was performed. Higher doses of IMN administered on an outpatient (at home) basis might have an even greater effect on reducing prenatal assessment ward stay, although this has yet to be investigated.
Further research is needed to confirm the merits of higher doses of IMN or other cervical ripening agents on an outpatient (at home) basis. In addition, value of information methodology based on data from our study could potentially be used to inform sample size and data requirements for future economic evaluations in this area.23,24 Care providers need to be more aware of the costs associated with delayed induction in post-term pregnancy and the benefits of at home induction in low risk women. One possible option would be to advise care providers about more optimal configurations of care for women to save costs, maximise bed space and possibly improve women’s experiences.
Disclosure of interests
Wellbeing did not play any role in the study design; collection, analysis, interpretation of the data; writing of the report; nor in the decision to submit the paper for publication. The authors have no other competing interests.
Contributions to authorship
Oya Eddama was responsible for the collection of data, analysis and interpretation of results, and writing of the paper. Stavros Petrou was responsible for the proposal and design of the study, as well as supervision of the work in this paper. Liz Schroeder was responsible for initial collection of cost data as well as aiding the final analysis. Shrikant Bollapragada was responsible for recruitment of patients, collection of data and execution of the clinical study. Fiona Mackenzie was responsible for the clinical proposal, design and supervision of the clinical study. John Norrie was responsible for the clinical study proposal, design and analysis of clinical data. Margaret Reid was responsible of clinical study proposal and design of the clinical study and for the interviews with the subjects as well as holding focus groups with the midwives. Jane Norman was responsible for the clinical study proposal, design, analysis of clinical data, supervision of the clinical work and was the Chief Investigator. All authors contributed to drafting the final manuscript.
Details of ethics approval
The clinical study was approved by the North Glasgow University NHS Trust ethics committee and written informed consent was gained prior to randomisation.
The IMOP Trial and the economic evaluation were funded by Wellbeing. The National Perinatal Epidemiology Unit is funded by the Department of Health, England. The views expressed by the authors do not necessarily reflect those of the funding bodies.