If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
School of Public Health and Preventive Medicine, Monash University, Melbourne, AustraliaWarwick Business School, Warwick University, Coventry, England, UK
World Health Organization Collaborating Centre for Global Women’s Health, Institute of Metabolism and Systems Research, University of Birmingham, Birmingham, England, UK
∗ Helena Teede and Zanfina Ademi contributed equally to this work and share joint senior authorship.
Helena Teede
Footnotes
∗ Helena Teede and Zanfina Ademi contributed equally to this work and share joint senior authorship.
Affiliations
School of Public Health and Preventive Medicine, Monash University, Melbourne, AustraliaMonash Health Endocrine and Diabetes Units, Monash Health, Melbourne, Australia
∗ Helena Teede and Zanfina Ademi contributed equally to this work and share joint senior authorship.
Zanfina Ademi
Correspondence
Correspondence: Zanfina Ademi, PhD, School of Public Health and Preventive Medicine, Faculty of Medicine Nursing and Health Sciences, Monash University, 553 St Kilda Rd, Melbourne, Australia 3004.
Although lifestyle intervention in pregnancy effectively reduces excess gestational weight gain and adverse pregnancy outcomes, limited research has yielded inconsistent findings on cost-effectiveness. It is also unclear which types of lifestyle interventions are the most cost-effective.
•
The meta-analysis on which this study is based has generated level 1 evidence of the efficacy of lifestyle interventions by intervention type in gestational weight gain and maternal and neonatal outcomes. The current cost-effectiveness study builds on this and shows that physical activity interventions appear cost saving and that diet and diet with physical activity interventions are likely to be cost-effective dependent on willingness-to-pay thresholds over a short time horizon and based on maternal pregnancy outcomes alone. Mixed interventions (lacking structured diet or physical activity intervention components) were not cost-effective. Scenario analysis, when diet-only, diet with physical activity, and physical activity-only interventions were analyzed, appeared cost saving. When neonatal intensive care unit costs were added to the model, all intervention types were cost saving except for mixed interventions.
•
These results favor implementing lifestyle interventions in pregnancy that incorporate structured diet and physical activity components at the population level.
Abstract
Objectives
Lifestyle interventions during pregnancy improve maternal and infant outcomes. We aimed to compare the cost-effectiveness of 4 antenatal lifestyle intervention types with standard care.
Methods
A decision tree model was constructed to compare lifestyle intervention effects from a novel meta-analysis. The target population was women with singleton pregnancies and births at more than 20 weeks’ gestation. Interventions were categorized as diet, diet with physical activity, physical activity, and mixed (lacking structured diet and, or, physical activity components). The outcome of interest was cost per case prevented (gestational diabetes, hypertensive disorders in pregnancy, cesarean birth) expressed as an incremental cost-effectiveness ratio (ICER) from the Australian public healthcare perspective. Scenario analyses were included for all structured interventions combined and by adding neonatal intensive care unit costs. Costs were estimated from published data and consultations with experts and updated to 2019 values. Discounting was not applied owing to the short time horizon.
Results
Physical activity interventions reduced adverse maternal events by 4.2% in the intervention group compared with standard care and could be cost saving. Diet and diet with physical activity interventions reduced events by 3.5% (ICER = A$4882) and 2.9% (ICER = A$2020), respectively. Mixed interventions did not reduce events and were dominated by standard care. In scenario analysis, all structured interventions combined and all interventions when including neonatal intensive care unit costs (except mixed) may be cost saving. Probabilistic sensitivity analysis showed that for physical activity and all structured interventions combined, the probability of being cost saving was 58% and 41%, respectively.
Conclusions
Governments can expect a good return on investment and cost savings when implementing effective lifestyle interventions population-wide.
Interventions to improve diet and physical activity behaviors in pregnancy can mitigate the impact of the obesogenic environment that affects women’s health worldwide.
International Weight Management in Pregnancy (i-WIP) Collaborative Group Effect of diet and physical activity based interventions in pregnancy on gestational weight gain and pregnancy outcomes: meta-analysis of individual participant data from randomised trials [published correction appears in BMJ. 2017 Aug 23;358:j3991].
Institute of Medicine and National Research Council Committee to Reexamine IOM Pregnancy Weight Guidelines Committee to re-examine IOM pregnancy weight guidelines.
in: Rasmussen K.M. Yaktine A.L. Weight Gain During Pregnancy: Reexamining the Guidelines. National Academies Press,
Washington, DC2009
In Australia, approximately 15% of pregnant women are expected to develop gestational diabetes mellitus (GDM), and these women are likely to have a higher incidence of hypertensive disease in pregnancy.
As well as affecting the health of women and their offspring, GDM and hypertensive disease in pregnancy have a high cost burden on healthcare systems, estimated at $1.3 billion for GDM and $1.03 billion for preeclampsia in the United States.
Large cost savings could potentially be generated by reducing the incidence of these 2 conditions, and recent evidence has shown that antenatal lifestyle interventions can limit gestational weight gain and prevent adverse events including GDM and cesarean birth.
International Weight Management in Pregnancy (i-WIP) Collaborative Group Effect of diet and physical activity based interventions in pregnancy on gestational weight gain and pregnancy outcomes: meta-analysis of individual participant data from randomised trials [published correction appears in BMJ. 2017 Aug 23;358:j3991].
Before recommending implementation and translation of these interventions, health economic analysis is needed to inform healthcare providers and policy makers.
In our previous systematic review of lifestyle interventions in pregnancy with an existing economic analysis, we were unable to determine overall cost-effectiveness because the few interventions evaluated were largely ineffective.
International Weight Management in Pregnancy (i-WIP) Collaborative Group Effect of diet and physical activity based interventions in pregnancy on gestational weight gain and pregnancy outcomes: meta-analysis of individual participant data from randomised trials [published correction appears in BMJ. 2017 Aug 23;358:j3991].
Results suggested that antenatal lifestyle interventions were likely to be cost-effective, with higher cost-effectiveness for higher body mass index (BMI) groups. In this modeling study, we were able to capture the overall impact of all interventions combined but not to investigate separate interventions to determine which yielded better value for money.
To be able to investigate different intervention types, we updated our previous systematic review and meta-analysis of randomized controlled trials of antenatal lifestyle interventions published in 2017, initially used in our modeling study.
International Weight Management in Pregnancy (i-WIP) Collaborative Group Effect of diet and physical activity based interventions in pregnancy on gestational weight gain and pregnancy outcomes: meta-analysis of individual participant data from randomised trials [published correction appears in BMJ. 2017 Aug 23;358:j3991].
Effect of antenatal lifestyle interventions including diet and physical activity on gestational weight gain and pregnancy outcomes: a systematic review and meta-analysis.
Effect of antenatal lifestyle interventions including diet and physical activity on gestational weight gain and pregnancy outcomes: a systematic review and meta-analysis.
were structured, typically presented by a dietitian and recommended a specific diet or types of foods. Physical activity interventions were structured, usually with 2 to 3 exercise classes per week over 15 to 26 weeks, delivered by a trained instructor. Diet with physical activity interventions included at least 1 structured diet and physical activity component. Mixed interventions lacked articulated structured diet and physical activity components (eg, unstructured, written lifestyle guidance that may include weight monitoring).
When implementing interventions in real-world settings, policy makers need reliable information on cost-effectiveness to appropriately allocate scarce resources.
Thus, understanding which types of interventions are more cost-effective is essential for policy decision making. In this study, we aimed to explore cost-effectiveness based on the data from the 117 studies recently captured in our systematic review
Effect of antenatal lifestyle interventions including diet and physical activity on gestational weight gain and pregnancy outcomes: a systematic review and meta-analysis.
across clinically prioritized maternal and neonatal outcomes and across these 4 different intervention types. This knowledge will underpin the implementation and scale-up of lifestyle interventions in pregnancy.
Methods
Overview of Outcomes and Model
We conducted a cost-effectiveness analysis using the decision tree model developed for our previous study.
: GDM, hypertensive disorders in pregnancy, and cesarean delivery. The incremental cost-effectiveness ratio (ICER) was defined as the cost per case prevented, with cases being GDM and hypertensive disorders in pregnancy and incorporating the effects of birth type (cesarean delivery and/or induction). We included 4 health states: developed GDM, developed a hypertensive disorder in pregnancy, developed both GDM and a hypertensive disorder in pregnancy, or developed neither GDM nor a hypertensive disorder in pregnancy. For each health state, 4 birthing outcomes were modeled: cesarean delivery, induction with vaginal birth, induction with cesarean delivery, and neither induction nor cesarean delivery. The decision tree model is shown in Appendix Figure 1 in Supplemental Materials found at https://doi.org/10.1016/j.jval.2021.07.013. Outcomes were considered from the perspective of the public healthcare system. The time horizon covered early pregnancy (trials generally started around the end of the first trimester) to birth (hospital stay post normal stay for the birth was not included); hence, discounting was not required.
Population
To define the baseline (standard care), a deidentified data set from the largest health service network in Australia was used that included routine maternity care outcomes such as GDM, hypertensive disorder in pregnancy, and type of birth, as outlined in our previous publication (no subgroups were excluded from the data set).
The population included in this data set can be considered representative of the general Australian population and includes high ethnic and socioeconomic diversity.
Pregnancy outcomes and insulin requirements in women with type 1 diabetes treated with continuous subcutaneous insulin infusion and multiple daily injections: cohort study.
Data were available for the years from 2009 to 2013 for singleton pregnancies with a birth of more than 20 weeks’ gestation. During this time, the GDM diagnostic criteria matched most closely to those used in the meta-analysis lifestyle intervention studies. BMI was categorized according to the World Health Organization recommendations.
), of which 54% of women were categorized as having a BMI less than 25 kg/m2, 27% had a BMI from 25 to 30 kg/m2, and 20% had a BMI greater than 30 kg/m2. Mean age was 29.7 years (SD 5.43), most women spoke English as their first language (70%), 56% of pregnancies were multiparous, and 83% of women did not smoke.
7.5% of women received a diagnosis of GDM. Rates of hypertensive disorder in pregnancy were 3.8%, and 0.5% of women had both conditions. Cesarean birthrates were 27.8%, and inductions were 21.9%. Mode of birth by health status has been previously published.
Approval for the analysis of the health network data set was attained from the Monash Health Human Research Ethics Committee (Approval14001Q).
Standard Care
Between 2009 and 2013, routine antenatal care received by women in the health network data set included generic written resources that were either mailed at booking or provided at the booking visit, usually at approximately 16 weeks’ gestation. It consisted of recommendations for occasional weight monitoring plus provision of information on food safety, hygiene, and national dietary guidelines. Women did not receive lifestyle or dietary interventions during pregnancy.
Lifestyle Intervention Effects
The mean intervention effect for each intervention type compared with standard care was obtained from a large, recently updated meta-analysis.
International Weight Management in Pregnancy (i-WIP) Collaborative Group Effect of diet and physical activity based interventions in pregnancy on gestational weight gain and pregnancy outcomes: meta-analysis of individual participant data from randomised trials [published correction appears in BMJ. 2017 Aug 23;358:j3991].
Effect of antenatal lifestyle interventions including diet and physical activity on gestational weight gain and pregnancy outcomes: a systematic review and meta-analysis.
Classification of intervention type was conducted independently by an experienced dietitian, exercise physiologist, and endocrinologist/public health physician. Interventions were categorized as structured diet, diet with physical activity where at least 1 component was structured, structured physical activity, and mixed (lacking structured diet and physical activity components). To account for the intervention effect, risk ratios were calculated from odds ratios as per our previous publication,
with odds ratios sourced from the recent meta-analysis (HT). Risk ratios, 95% confidence intervals (CIs), and distributions are presented in Table 1 for health states and birth outcomes.
Table 1Key input parameters for effects and costs for GDM, HDP, both GDM and HDP, cesarean delivery, and induction, by intervention type, including parameter variation and distribution.
Data for both GDM and HDP were modeled on GDM outcomes because this outcome was not measured in the meta-analysis.
0.814
0.731
0.904
Lognormal
Cesarean delivery
0.956
0.910
1.007
Lognormal
Induction
1
Costs A$
Base
−30%
30%
Antenatal GDM costs
1055
738
1371
Gamma
Antenatal GDM and HDP costs
2781
1947
3616
Gamma
Antenatal HDP costs
1923
1346
2500
Gamma
Intervention costs
228
160
296
Gamma
Vaginal birth costs
5812
4068
7555
Gamma
Cesarean delivery costs
11 416
7992
14 841
Gamma
Induction and vaginal birth
7846
5492
10 200
Gamma
Induction only
2034
1424
2645
Gamma
Note. Risk ratios for “both GDM and HDP” were set to those for GDM. Risk ratios were not available for induction. A merged group was created to capture all structured diet, physical activity, and diet with physical activity interventions.
GDM indicates gestational diabetes mellitus; HDP, hypertensive disorders in pregnancy.
∗ Data for both GDM and HDP were modeled on GDM outcomes because this outcome was not measured in the meta-analysis.
To estimate average intervention costs, we sourced information on intervention components from the studies in the meta-analysis where GDM, hypertensive disorders in pregnancy, and/or cesarean delivery outcomes were included. Data were extracted for the following: (1) practitioner type, (2) contact minutes, and (3) whether delivered individually or in groups. Three assumptions were made to account for missing information. First, if delivered by group but group size was not stated, it was assumed as 10 in the base case. Group sizes of 6 or 15 were investigated in scenario analyses. Second, cost per hour for the health professional delivering the intervention was obtained from Medicare Benefits Schedule (MBS) data
(MBS data provide costs of health services funded by the Australian Government in the context of a universal, freely accessible public health system). Costs for fitness instructors were not included in the MBS schedule and were sourced from the Fitness Australia Award Pay Guide.
Third, where session length was not available, we assumed 60 minutes for initial individual visits, 30 minutes for subsequent individual visits, and 45 minutes for group classes. We estimated home visits at 90 minutes to include travel time. Follow-up phone calls were estimated at 20 minutes. Despite these assumptions, there was insufficient information to calculate costs in 10 interventions.
The Treatment of Obese Pregnant Women (TOP) study: a randomized controlled trial of the effect of physical activity intervention assessed by pedometer with or without dietary intervention in obese pregnant women.
Dietary approaches to stop hypertension diet and activity to limit gestational weight: maternal offspring metabolics family intervention trial, a technology enhanced randomized trial.
A Mediterranean diet with additional extra virgin olive oil and pistachios reduces the incidence of gestational diabetes mellitus (GDM): a randomized controlled trial: the St. Carlos GDM prevention study.
Costs for patient pathways and unit costs are presented in the Appendix Tables 1 to 4 in Supplemental Materials found at https://doi.org/10.1016/j.jval.2021.07.013. Average intervention costs by category were A$168 for diet, A$187 for diet with physical activity, A$217 for physical activity, A$184 for mixed, and A$198 for all interventions combined. Interquartile range and minimum and maximum costs by intervention type are presented in Appendix Table 5 in Supplemental Materials found at https://doi.org/10.1016/j.jval.2021.07.013.
Public Healthcare Costs
Costs of patient care in the public health system were outlined in our previous publication.
In brief, we developed patient pathways for GDM, hypertensive disorders in pregnancy, and cesarean delivery to approximate the type and nature of services that women were likely to access. Patient pathways were developed using clinical guidelines and policies, published articles, and multidisciplinary expert opinion. Antenatal care, associated costs, and medication costs were sourced from the MBS
National Hospital Cost Data Collection, Public Hospitals Cost Report, Round 20 (Financial Year 2015-16). Independent hospital pricing authority (IHPA).
Input parameters for costs are additionally presented in Table 1 (note that neonatal intensive care unit [NICU] costs were not included in the base-case analysis but are considered in the third scenario analysis).
Scenario Analysis
Scenario analysis 1
Group size was an important indicator of intervention costs in physical activity and mixed interventions; the group size, however, was not stated in 26 of the 43 interventions with physical activity components. In the base case, a class size of 10 was used if not otherwise stated. In this scenario analysis, 2 further models were conducted assuming 6 persons per group and 15 persons per group.
Effect of antenatal lifestyle interventions including diet and physical activity on gestational weight gain and pregnancy outcomes: a systematic review and meta-analysis.
a sensitivity analysis was conducted in which data for structured interventions across diet, diet with physical activity, and physical activity interventions were analyzed together. In scenario 2, we used the risk ratios from this combined analysis. Risk ratios for the overall structured intervention data were as follows: GDM (and both GDM and hypertensive disorders in pregnancy) (0.670 [95% CI 0.582-0.765]), hypertensive disorders in pregnancy (0.737 [95% CI 0.600-0.889]), and cesarean birth (0.929 [95% CI 0.856-1.008]). Average intervention costs across these structured intervention categories over the time horizon were A$203.
Scenario analysis 3
In the base case, we included costs for maternal outcomes (GDM, hypertensive disorders of pregnancy, and cesarean delivery). A third scenario analysis was conducted to measure the effect of admission to NICU, which was not considered in the base-case analysis. Total average weighted NICU costs were estimated from 8 Diagnosis-Related Group codes relating to neonatal costs
National Hospital Cost Data Collection, Public Hospitals Cost Report, Round 20 (Financial Year 2015-16). Independent hospital pricing authority (IHPA).
(see Appendi Table 6 in Supplemental Materials found at https://doi.org/10.1016/j.jval.2021.07.013 for further information on the codes). The percentage differences between intervention and standard care groups for NICU costs were calculated from the actual percentage of events in the meta-analysis. The cost of NICU was estimated at A$76 621 per admission. NICU costs were added to baseline costs and ICERs calculated.
Sensitivity Analysis
Deterministic sensitivity analysis
One-way sensitivity analysis was used to estimate the impact of uncertainty around effect and cost parameters for analyzing structured interventions together. Risk ratios were varied by 95% CIs and costs by ±30%.
To explore combined parameter uncertainty, probabilistic sensitivity analysis was conducted (10 000 iterations) and displayed on a cost-effectiveness plane. Risk ratios were varied by 95% CIs and cost by ±30%. In line with our previous study,
we used gamma distributions for cost inputs and lognormal distributions for effects.
Statistical Analysis
Health-center data were analyzed using Stata version 15 (StataCorp LLC, College Station, TX). Decision tree models were built in Microsoft Excel (Microsoft Corporation, Redmond, WA). ICERs were calculated, representing cost per case (GDM, hypertensive disorder in pregnancy, or both) prevented and accounting for birth type (vaginal, cesarean, induction). @RISK software version 7.5 in Excel was used for probabilistic sensitivity analyses.
Results
Comparison of 4 Intervention Types on Key Maternal Outcomes
Effect and cost differences and ICERs were calculated for each intervention type, as shown in Table 2. For physical activity interventions, there was a 4.2% difference in effects between groups, and the intervention was likely to be cost saving (58.4% of pairs lay in the southeast quadrant suggesting higher effects and lower costs). In the diet-only and diet with physical activity interventions, there were 3.5% and 2.9% differences in effects between groups, respectively, with cost differences of A$169 and A$59 per person, respectively. ICERs were A$4882 for diet and A$2020 for diet with physical activity interventions (northeast quadrant). Mixed interventions had limited efficacy between intervention and standard care groups (−0.07%), with a cost difference of A$182 per person and an ICER of −A$27 020, thus dominated by standard care. Comparisons between the intervention groups are presented on a cost-effectiveness plane in Figure 1.
Table 2Effects (percentage of complications avoided), costs, and ICERs with 95% CI from PSA comparing standard care with intervention by intervention category for maternal outcomes.
Because of the very small effect size causing a large ICER, it was not possible to run probability sensitivity analysis and obtain CIs or quadrant percentages for these estimates.
∗ Because of the very small effect size causing a large ICER, it was not possible to run probability sensitivity analysis and obtain CIs or quadrant percentages for these estimates.
We estimated 10 participants per class in the base case where class size was not stated and 6 per class and 15 per class size in this scenario analysis. Results are shown in Appendix Table 7 in Supplemental Materials found at https://doi.org/10.1016/j.jval.2021.07.013. Changing the class size for the physical activity interventions had the largest effect on the ICER; physical activity interventions, however, remained cost saving. Diet with physical activity classes varied but remained in the northeast quadrant. Diet-only and mixed interventions were unaffected (as they did not include physical activity classes).
Scenario analysis 2
In scenario 2, data for structured interventions across diet, diet with physical activity, and physical activity interventions were analyzed together. Effects here were 8.2% compared with 11.8% for standard care with a cost difference of 3.6% (95% CI: 2.73, 4.49). Costs were A$8240 for the intervention group and A$8248 for the standard care with cost saving of A$8 (95% CI: −911, 601). The ICER was dominant (cost-saving) at A$ −217 (95% CI: −25 583, 16 732). Forty-one percent of values in the PSA fell in the North-East quadrant and 59% in the South-East quadrant.
Scenario analysis 3
To understand the additional impact of NICU admission on the ICER, we added NICU costs for intervention and standard care groups, whereas all other parameters remained the same. When NICU costs were added, all ICERs were in the southeast quadrant (cost saving), except for the mixed interventions that were in the northwest quadrant (dominated) as shown in Table 3.
Table 3Percentage of NICU admissions, costs of NICU, and total costs by intervention and standard care groups and ICERs to add the costs and effects of NICU into the model.
Deterministic (one-way) sensitivity analysis was conducted on all interventions combined (including mixed). The tornado graph is presented in Figure 2. The most influential parameters in the one-way sensitivity analysis were risk ratios and costs of cesarean delivery, intervention costs, GDM and hypertensive disorders of pregnancy risk ratios, and vaginal birth costs. Tornado graphs for diet, diet and physical activity, and physical activity are shown in Appendix Figures 2 to 4 in Supplemental Materials found at https://doi.org/10.1016/j.jval.2021.07.013. The cesarean delivery risk ratio had the highest effect on each of these intervention types.
Figure 2Summary of the deterministic sensitivity analysis results presented in a tornado chart. Blue bars represent lower limit values. Gray bars represent upper limit values.
Probabilistic sensitivity analyses were conducted for each intervention type separately (diet, diet and physical activity, physical activity) and all interventions combined. The cost-effectiveness plane with 10 000 iterations using gamma distributions for costs and log normal distributions for effects is presented in Table 2 and Figure 3. In addition, cost-effectiveness acceptability curves were presented in Appendix Figures 5 to 8 in Supplemental Materials found at https://doi.org/10.1016/j.jval.2021.07.013. For physical activity interventions, the probability of being cost saving was 54% compared with 28% for all interventions combined. For scenario 2, the probabilistic sensitivity analysis for all structured interventions (diet, diet and physical activity, physical activity) combined indicated a 95% CI of −A$29 284 to A$23 886, with 50.8% of values lying in the northeast quadrant and 49.2% of values in the southeast quadrant.
Figure 3Cost-effectiveness plane demonstrating the probability of cost-effectiveness with 10 000 iterations using gamma distributions for costs and lognormal distributions for effects of diet, physical activity, and diet plus physical activity compared with standard care.
There is now a clear, level 1 evidence of the efficacy of lifestyle interventions in improving gestational weight gain and reducing adverse pregnancy outcomes.
Effect of antenatal lifestyle interventions including diet and physical activity on gestational weight gain and pregnancy outcomes: a systematic review and meta-analysis.
Here, we modeled the cost-effectiveness of antenatal lifestyle interventions over a short time horizon in terms of preventing common maternal adverse outcomes (GDM, hypertensive disorders in pregnancy, and cesarean delivery). In the base case, structured physical activity interventions may be cost saving, and structured diet and “diet with physical activity” interventions may be cost-effective, dependent on willingness-to-pay thresholds. Mixed interventions were not cost-effective, with the ICER situated in the northwest quadrant (less effective, more costly). In scenario analysis, when structured interventions across diet, diet with physical activity, and physical activity interventions were analyzed together, these interventions appeared cost saving. When NICU costs were added to the model, all intervention types appeared cost saving except for mixed interventions (which were not cost-effective). These findings suggest that lifestyle interventions in pregnancy including structured diet, structured physical activity or diet, and physical activity with at least 1 component being structured are likely to be cost saving or cost-effective.
Previous cost-effectiveness studies on antenatal lifestyle interventions have been largely negative, mainly because of inadequate sample size and failure to demonstrate individual study intervention efficacy on pregnancy outcomes.
International Weight Management in Pregnancy (i-WIP) Collaborative Group Effect of diet and physical activity based interventions in pregnancy on gestational weight gain and pregnancy outcomes: meta-analysis of individual participant data from randomised trials [published correction appears in BMJ. 2017 Aug 23;358:j3991].
International Weight Management in Pregnancy (i-WIP) Collaborative Group Effect of diet and physical activity based interventions in pregnancy on gestational weight gain and pregnancy outcomes: meta-analysis of individual participant data from randomised trials [published correction appears in BMJ. 2017 Aug 23;358:j3991].
Effect of antenatal lifestyle interventions including diet and physical activity on gestational weight gain and pregnancy outcomes: a systematic review and meta-analysis.
and comparing ICERs across a range of intervention types using the average of costing from individual interventions. The ICER for all interventions was A$1470 in our previous analysis
and A$3855 in the current analysis; differences were due to updated risk ratios in the meta-analyses and more reliable intervention cost estimates (we modelled all interventions and compared 4 intervention types over a larger pool of interventions). These results strongly support the cost-effectiveness of antenatal lifestyle interventions applying structured diet, physical activity, or diet and physical activity interventions.
The deterministic sensitivity analysis suggested that cesarean delivery costs and effects had a large influence on cost-effectiveness in the model. Because cesarean delivery was experienced by 28% of women (similar to approximately 30% of births Australia-wide
), costs for cesarean deliveries were likely to have a significant impact. We also found that there was a significant effect on cesarean delivery rates from lifestyle interventions during pregnancy in the meta-analysis results.
Effect of antenatal lifestyle interventions including diet and physical activity on gestational weight gain and pregnancy outcomes: a systematic review and meta-analysis.
Intervention costs also had a strong influence on the model, suggesting that the cost of delivering a lifestyle intervention at a population level needs to be carefully considered. Risk ratios for GDM and hypertensive disorders in pregnancy contributed to uncertainty, as did hospital costs for vaginal births. The probabilistic sensitivity analyses showed that CIs were wide; these results reflect highly variable costs in the real world.
A key consideration in this analysis was that the model was limited to the clinical prioritized maternal outcomes, excluding neonatal outcomes. We have included the neonatal outcome of NICU in scenario analysis; however, future models may be able to include outcomes specific to the different intervention categories. For example, structured diet has the greatest impact on gestational weight gain and GDM, preterm delivery, large for gestational age, NICU admission, and total maternal and neonatal adverse outcomes.
Effect of antenatal lifestyle interventions including diet and physical activity on gestational weight gain and pregnancy outcomes: a systematic review and meta-analysis.
Diet also reduced gestational weight gain by approximately 20%, which was more than physical activity alone. The effect of diet may influence postpartum weight retention, long-term obesity, and noncommunicable disease risk, with additional anticipated cost benefits. Finally, the model did not include broader maternal or neonatal benefits in pregnancy or outcomes over the longer term, such as prevention of maternal and childhood obesity and noncommunicable diseases. Including these approaches may improve the cost-effectiveness demonstrated here, as our results are likely to have underestimated the cost benefits of antenatal lifestyle interventions due to not having included these factors. Once long-term efficacy data are available, these outcomes would be important to include in future studies.
Implementation and scale-up of the most effective lifestyle programs into routine healthcare have now been funded in Australia. The authorship team has influenced national guidelines for healthy lifestyle support in routine care,
The healthy pregnancy service to optimise excess gestational weight gain for women with obesity: a qualitative study of health professionals’ perspectives.
Early results suggest high efficacy and strong support from clinicians and women and high feasibility, with publications currently under review.
Strengths and Limitations
The strengths of this study were that the intervention effect was based on a large meta-analysis, intervention costs were estimated from more than 70 interventions, and estimates of the prevalence of complications were from a large health network data set. Nevertheless, a limitation was that interventions may differ substantially both within and across categories. Further research from our team is ongoing to determine which intervention components are most effective, which intervention types work for whom, and in what circumstances. Notably, owing to the nature of the aggregated study data in the meta-analysis, subanalysis by BMI category was not feasible. Subgroup analysis by BMI category was undertaken in our previous modeling publication where we found that interventions were more likely to be cost saving or cost-effective at higher weight.
Because actual costs for each intervention were not available, we estimated costs based on information supplied in each article on practitioner time in intervention delivery; however, 10 interventions could not be assessed because of insufficient information. Costs and effects may be less reliable for diet interventions because the effect estimates were based on only 9 interventions and costs on only 6 interventions. Costing patient pathways for each complication type was based on published data, guidelines and clinical expertise, but was challenging because of the large range of possible costs associated (for instance, a case of hypertensive disorder in pregnancy may result in an early induction with little extra cost or in prolonged costly hospital stays). Nonetheless, the cost estimates reflect real-world clinical scenarios. Finally, the model did not include broader benefits over the longer term, such as neonatal outcomes or maternal obesity and noncommunicable diseases, which are important for future studies.
Conclusion
Structured antenatal diet, physical activity, or diet with physical activity interventions appear cost saving or cost-effective based on willingness-to-pay thresholds when only clinically prioritized, short-term maternal outcomes are considered. Notably, we leveraged efficacy data on lifestyle interventions from a recent systematic review and meta-analysis of 117 intervention studies and estimated cost of each intervention. Other neonatal and longer-term benefits were not included in the model, suggesting that cost-effectiveness is likely to have been underestimated. These results support the implementation and scale-up of lifestyle interventions in pregnancy at the population level to improve the health of women and the next generation. This work is now extensively funded by the Australian Government and underway.
Article and Author Information
Author Contributions:Concept and design: Bailey, Skouteris, Hill, Teede, Ademi
Acquisition of data: Bailey, Harrison, Thangaratinam, Teede, Ademi
Analysis and interpretation of data: Bailey, Teede, Ademi
Drafting of the manuscript: Bailey, Teede, Ademi
Critical revision of manuscript for important intellectual content: Bailey, Skouteris, Harrison, Hill, Thangaratinam, Teede, Ademi
Statistical analysis: Bailey, Teede, Ademi
Provision of study materials or patients: Harrison, Teede
Obtainingfunding: Skouteris, Teede, Ademi
Administrative, technical, or logisticsupport: Harrison, Thangaratinam, Teede
Supervision: Hill, Skouteris, Teede, Ademi
Conflict of Interest Disclosures: Dr Bailey reported receiving the Australian Government Research Training Program (RTP) Scholarship during the conduct of this study. Drs Hill and Teede reported receiving grants from the National Health and Medical Research Council during the conduct of the study. No other disclosures were reported.
Funding/Support: Funding for this research has been provided from the Australian Government’s Medical Research Future Fund (MRFF). The MRFF provides funding to support health and medical research and innovation, with the objective of improving the health and well-being of Australians. MRFF funding has been provided to The Australian Prevention Partnership Centre under the MRFF Boosting Preventive Health Research Program. Further information on the MRFF is available at www.health.gov.au/mrff. The UK National Institute for Health Research supported the iWIP research.
Role of the Funder/Sponsor: The funder had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.
Ethics: No ethics approval was required.
Acknowledgment
We acknowledge Assistant Professor Emily Callander, Professor Danny Liew, Dr Catherine Keating, Clara Marquina, Dr Melanie Llloyd and the Monash Outcome Research and Health Economics team (MORE) for their support in completing this project.
International Weight Management in Pregnancy (i-WIP) Collaborative Group
Effect of diet and physical activity based interventions in pregnancy on gestational weight gain and pregnancy outcomes: meta-analysis of individual participant data from randomised trials [published correction appears in BMJ. 2017 Aug 23;358:j3991].
Effect of antenatal lifestyle interventions including diet and physical activity on gestational weight gain and pregnancy outcomes: a systematic review and meta-analysis.
Pregnancy outcomes and insulin requirements in women with type 1 diabetes treated with continuous subcutaneous insulin infusion and multiple daily injections: cohort study.
The Treatment of Obese Pregnant Women (TOP) study: a randomized controlled trial of the effect of physical activity intervention assessed by pedometer with or without dietary intervention in obese pregnant women.
Dietary approaches to stop hypertension diet and activity to limit gestational weight: maternal offspring metabolics family intervention trial, a technology enhanced randomized trial.
A Mediterranean diet with additional extra virgin olive oil and pistachios reduces the incidence of gestational diabetes mellitus (GDM): a randomized controlled trial: the St. Carlos GDM prevention study.
National Hospital Cost Data Collection, Public Hospitals Cost Report, Round 20 (Financial Year 2015-16). Independent hospital pricing authority (IHPA).
The healthy pregnancy service to optimise excess gestational weight gain for women with obesity: a qualitative study of health professionals’ perspectives.
01655-7&id=gr1.jpg){kind=link}
01655-7&id=gr2.jpg){kind=link}
01655-7&id=gr3.jpg){kind=link}