Results
Health outcomes, resource use and costs are summarised in Table 1 by method of IOL. Most women in the balloon and PG group were nulliparous (72.7% and 67.1%, respectively) and birthed vaginally (67.8% and 73.6% respectively), with more women in the PG group starting IOL with a more favourable cervix (MBS>3) compared to the balloon group (56.2% vs 47.7%). Health utility index scores showed lower health scores for nulliparous (0.726 balloon vs 0.694 PG group) than for parous women (0.80 balloon vs 0.834 PG groups) and higher health scores for vaginal birth (0.804 unassisted, 0.705 instrumental) compared to CS (0.643). Health utility scores were marginally higher for women starting IOL with a less favourable (MBS≤3 compared to more favourable cervix (0.744 vs 0.738). Balloon catheters cost $49 and the mean cost of PG cervical ripening was $105. Hourly cost of care in the antenatal ward was $51, $259 in the birth suite, $48 in the postnatal ward and $81-$125 in the neonatal nurseries.
Deterministic model analyses showed costs in the balloon group (n=205) and PG group (n=243), were $9,765 versus $10,272 per QALY gained, retrospectively (see Table 2). The balloon group absolutely dominated the PG group with lower mean costs ($7,294 vs $7,585) and higher mean effectiveness (0.75 vs 0.74 QALYs), resulting in a negative ICER which indicates cost savings of $34,193 per quality-adjusted life year gained if this method of induction was chosen. The net monetary benefit (NMNB) of outpatient-balloon IOL was $30,054 compared to $29,338 in the PG group.
In probabilistic sensitivity analyses (PSA) using Monte Carlo simulation, the balloon group had a higher probability of being cost-effective at each WTP threshold tested (see Figure 1). At a WTP of $50,000 the likelihood of outpatient-balloon IOL being cost-effective is 55.25% compared to inpatient-PG IOL (44.75%). Figure 2 shows the scatter plot on the incremental cost-effectiveness plane for both methods. The eclipse represents the 95% confidence interval of all simulated pairs of incremental cost-effectiveness from the 10,000 simulations. A total of 34.2% of all simulations fell in the south-east quadrant representing possible results in which outpatient-balloon IOL was less costly and more effective than inpatient-PG IOL. A further 17.8% of simulations fell in the north-east quadrant below the $50,000 WTP threshold, indicating higher costs and higher benefits of outpatient-balloon IOL that we would be willing-to-pay for each additional QALY (below $50,000). Another 3.3% of simulations in the south-west quadrant indicate lower costs and lower benefits of outpatient-balloon IOL below the WTP threshold. Overall, the area to the right of the WTP threshold line is the number of simulations (34.2% + 17.8% + 3.3% = 55.3%) in which outpatient-balloon IOL represents the better value for money, and the area to the left represents simulations in which inpatient-PG IOL is cost-effective (44.7%).
Extracts of results from 1-way sensitivity analyses are illustrated in the Supplementary material, Table S1. When the mean operating costs per CS with baseline value of $3,447 were varied from $1,000-$15,000, outpatient-balloon IOL had a higher NMB for each tested value and dominated inpatient-PG IOL up to a cost of $8,442. Furthermore, varying the mean hourly cost in antenatal ward of $51 from $0-$150 also showed consistently higher NMB for balloon induction which had lower costs and higher effectiveness (absolute dominance) for values between $28 to $150. One-way sensitivity analysis of mean hourly costs in birth suite with baseline value of $259 was tested for values between $100-$500 and resulted in higher NMB for each value with lower costs and higher effectiveness in the balloon group for values up to $450. The tested ranges of mean costs of stay in postnatal ward (baseline value $48, range $0 - $100), ICN (baseline value $125, range $50 - $200) and SCN (baseline value $81, range $0 - $300) resulted in absolute dominance (lower cost, higher effectiveness) of outpatient-balloon IOL with higher NMB ($27,399 - $32,505) compared to inpatient-PG IOL (NMB from $26,757 - $31,721).
Subgroup analyses by parity also indicated a higher likelihood of outpatient-balloon IOL being cost-effective for nulliparous women (64.5%) and inpatient-PG IOL being cost-effective for parous women (66.9%). The NMB for inpatient-PG IOL in parous women was the highest out of all types of analyses (see Table 2) with $36,969 compared to balloon induction in parous women with $34,210 and very low mean costs for parous women in both groups with $4,903 for PG and $5,659 in the balloon group. Subgroup analyses by cervix favourability showed higher mean costs for both methods of IOL for women with a more favourable cervix (MBS >3) compared to a less favourable cervix (MBS ≤3), and higher probability (59.1%) of outpatient-balloon IOL being cost-effective for women with a more favourable cervix (NMB $30,599 vs $27,904). Women with a less favourable cervix had lower mean costs by $759 and higher utility scores by 0.026 in the inpatient-PG IOL with 56.3% of simulations indicating cost-effectiveness.