(C) PLOS One
This story was originally published by PLOS One and is unaltered.
. . . . . . . . . .
The cost-effectiveness of preventing, diagnosing, and treating postpartum haemorrhage: A systematic review of economic evaluations [1]
['Joshua F. Ginnane', 'Maternal', 'Child', 'Adolescent Health Program', 'Burnet Institute', 'Melbourne', 'Samia Aziz', 'School Of Public Health', 'Preventive Medicine', 'Monash University']
Date: 2024-10
In this systematic review, we extracted, critically appraised, and summarised the cost-effectiveness evidence from 56 studies across 16 different interventions for the prevention, diagnosis, and treatment of PPH. Both the use of tranexamic acid as part of PPH treatment, and the use of comprehensive PPH bundles for prevention, diagnosis, and treatment have supportive cost-effectiveness evidence across a range of settings. More studies utilizing best practice principles are required to make stronger conclusions on which interventions provide the best value. Several high-priority interventions recommended by World Health Organization (WHO) such as administering additional uterotonics, non-pneumatic anti-shock garment, or uterine balloon tamponade (UBT) for PPH management require robust economic evaluations across high-, middle-, and low-resource settings.
From 3,993 citations, 56 studies were included: 33 studies of preventative interventions, 1 study assessed a diagnostic method, 17 studies of treatment interventions, 1 study comparing prevention and treatment, and 4 studies assessed care bundles. Twenty-four studies were conducted in high-income countries, 22 in upper or lower middle-income countries, 3 in low-income countries, and 7 studies involved countries of multiple income levels. Study settings, methods, and findings varied considerably. Interventions with the most consistent findings were the use of tranexamic acid for PPH treatment and using care bundles. In both cases, multiple studies predicted these interventions would either result in better health outcomes and cost savings, or better health outcomes at acceptable costs. Limitations for this review include that no ideal setting was chosen, and therefore, a transferability assessment was not undertaken. In addition, some sources of study uncertainty, such as effectiveness parameters, were interrogated to a greater degree than other sources of uncertainty.
This systematic review was prospectively registered on PROSPERO (CRD42023438424). We searched Medline, Embase, NHS Economic Evaluation Database (NHS EED), EconLit, CINAHL, Emcare, Web of Science, and Global Index Medicus between 22 June 2023 and 11 July 2024 with no date or language limitations. Full economic evaluations of any postpartum intervention for prevention, detection, or management of PPH were eligible. Study screening, data extraction, and quality assessments (using the CHEC-E tool) were undertaken independently by at least 2 reviewers. We developed narrative syntheses of available evidence for each intervention.
Considering the cost-effectiveness of interventions for PPH prevention, detection, and treatment is important for 2 reasons. Firstly, PPH disproportionately affects women in LMICs—in these countries, financial pressures on the health system mean that difficult decisions are taken around what interventions can be offered [ 4 ]. Cost-effectiveness analyses are therefore especially useful to national decision-makers to ensure optimal health impacts for available budget. Secondly, the 2023 WHO Roadmap to Combat Postpartum Haemorrhage has foreshadowed new consolidated WHO PPH guidelines [ 4 ]. Developing these guidelines involves explicit consideration of economic evidence, to assess the cost-effectiveness and resource requirements of candidate interventions. Previous systematic reviews on the cost-effectiveness of PPH interventions have solely focussed on the use of uterotonics for PPH prevention [ 11 ], or the use of UBT [ 12 ] or tranexamic acid for PPH treatment [ 13 ]. This systematic review thus aimed to broaden the evidence available from previous reviews by identifying, assessing, and synthesising all available evidence from economic evaluations of any postpartum intervention for PPH prevention, diagnosis, or treatment.
Deaths from PPH have largely been eliminated in high-income countries, due to coordinated interventions aimed at preventing, detecting, and treating PPH [ 4 ]. Preventative interventions include the use of prophylactic uterotonics such as oxytocin, carbetocin, ergometrine/methylergometrine, or misoprostol for all women immediately after birth. Alongside uterotonics, tranexamic acid can also be administered to further reduce the risk of excessive blood loss and the requirement for blood transfusion [ 7 ]. To diagnose PPH promptly, assessment of uterine tone and the measurement of blood loss postdelivery is recommended. Blood loss is either visually estimated or measured through collecting blood in a calibrated drape, weighing blood-soaked materials or the use of more sophisticated colorimetric systems [ 8 , 9 ]. If PPH occurs despite preventative measures, treatment should be commenced as soon as possible—this includes uterine massage, additional uterotonics, tranexamic acid, and intravenous fluids. If bleeding continues (i.e., refractory PPH), the use of devices such as uterine balloon tamponade (UBT) or a non-pneumatic antishock garment, compressive measures, and the administration of blood products may be required [ 10 ]. More invasive interventions may be required if bleeding continues. These include surgical treatments such as uterine compression sutures, uterine artery ligation or embolization, or hysterectomy [ 10 ]. At the health system level, implementing PPH-related care protocols, as well as ensuring adequate training and supervision for health workers, is key to ending PPH-related morbidity and mortality.
Postpartum haemorrhage (PPH) is a time-critical obstetric emergency, defined by the World Health Organization (WHO) as postpartum blood loss of more than 500 ml, regardless of mode of birth [ 1 ]. PPH affects approximately 6% of women giving birth and is the most common direct cause of maternal mortality, responsible for an estimated 19.7% of maternal deaths [ 2 , 3 ]. The incidence and resulting maternal mortality caused by PPH are disproportionately concentrated in low- and middle-income countries (LMICs) [ 4 ]. In addition to the health considerations, PPH also burdens health systems financially. Births complicated by PPH incur costs 21% to 309% higher than births without complications in LMICs [ 5 ]. Since 2015, progress on reducing the global maternal mortality ratio (MMR) appears to have stalled. It seems increasingly unlikely that the Sustainable Development Goals target 3.1—to reduce the global MMR to less than 70 per 100,000 live births by 2030—will be met [ 4 , 6 ]. The 2020 MMR estimates indicate that 223 maternal deaths occur per 100,000 live births worldwide [ 6 ]. Renewed efforts are required to address the underlying drivers of preventable pregnancy-related deaths, including the most common cause, PPH.
Studies were grouped for analysis by the characteristics and goal of the intervention, namely PPH prevention, diagnosis, or treatment. Due to heterogeneity of study designs within and between interventions, a narrative synthesis approach was used. A transferability assessment was not completed as no ideal setting was selected to compare to. To identify evidence gaps, we mapped included studies to current WHO recommendations. All guidance from the WHO recommendations for the prevention and treatment of PPH from 2012 were reviewed [ 1 ], in addition to the various individual updates that WHO have published to these recommendations up until the end of 2023 [ 23 – 29 ].
Costs were extracted unadjusted, in the year and currency stated by the study. For ease of comparison, all costs were converted to 2023 United States Dollars (USD) using an online tool developed by the Campbell and Cochrane Economics Methods Group (CCEMG) and the Evidence for Policy and Practice Information and Coordinating Centre (EPPI-Centre) [ 19 ]. This tool completes a two-stage cost conversion, first adjusting for the cost-year using a gross domestic product (GDP) deflator index, and then a currency conversion using purchasing power parities (PPPs) for GDP [ 20 ]. Quality assessments of included studies were completed independently by 2 reviewers (JFG, SA, SS) using the Consensus on Health Economics Criteria extended list (CHEC-E; Table A in S4 Appendix ) as this is suitable to evaluate economic evaluations based on clinical trials or modelling studies [ 14 , 21 , 22 ]. Studies scoring 0% to 49% were considered “low quality,” 50% to 74% as “moderate quality,” and >75% as “high quality.” Disagreements on data extraction or CHEC-E assessment were resolved through discussion or consulting other reviewers.
Cost-effectiveness analysis is highly dependent on the effectiveness estimate used; the accuracy of these estimates is thus a possible source of bias [ 18 ]. For each study, we extracted the effectiveness data used and compared it to an up-to-date effectiveness estimate from the corresponding WHO recommendation and systematic reviews of that intervention. When an included study based its cost-effectiveness calculations on an effectiveness estimate that differed significantly (outside of 95% confidence intervals) from an effectiveness estimate found in the literature, we flagged this as at risk of bias.
The search strategy was developed with an expert librarian and run on Medline, Embase, NHS Economic Evaluation Database, EconLit, CINAHL, Emcare, Web of Science, and Global Index Medicus ( S2 Appendix ). A scoping review of 923 economic evaluations of maternal health interventions was also searched for relevant articles [ 16 ]. Resulting citations were uploaded into Covidence software where at least 2 authors (JFG, SA, SS, CA, KE, MDS) independently screened citations by title/abstract level. Full texts of potentially eligible studies were reviewed by 2 reviewers (JFG, SA, SS, CA, KE, MDS). We recorded reasons for exclusion of any full texts. Disputes were resolved through discussion or input from other reviewers.
Full economic studies that have evaluated the cost-effectiveness, cost-utility, or cost-benefit of any method of PPH prevention, diagnosis, or treatment delivered in the postpartum period (from time of birth until 42 days postpartum) were included. We defined full economic evaluations as any evaluation that considered both health consequences and cost consequences of an intervention against a comparator in the single analysis. We considered any such intervention to be eligible, regardless of its effectiveness, or whether it is recommended by WHO or not. Effectiveness studies that included a full economic evaluation such as randomised trials, non-randomised interventional studies, or observational studies (cohort, case-control, cross sectional designs) were also eligible. Partial economic evaluations—those which did not consider both economic and health outcomes compared to a comparator—were excluded. Conference abstracts, protocols, grey literature, and reviews of existing evidence were excluded. There were no limitations based on date or language.
For this review, we followed guidelines from the Expert Review of Pharmacoeconomics and Outcomes Research [ 14 ] and reported findings in line with the Preferred Reporting Items for Systematic reviews and Meta-analyses (PRISMA) 2020 statement (Table A in S1 Appendix ) [ 15 ]. We prospectively registered the review protocol on PROSPERO (CRD42023438424). All included articles have been previously published, and ethics approval was not required.
Two studies evaluated bundles combining early detection and treatment interventions [ 91 , 93 ]. A study in Wales reported that their bundle—comprising universal risk assessments for PPH, early PPH identification through quantitative measurement, mandatory multidisciplinary team involvement at certain blood loss thresholds, and point-of-care coagulation testing after 1,000 ml blood loss—modestly reduced the proportion of PPH cases progressing to massive PPH, at a cost of $29.16 per patient with PPH > 1,000 ml [ 91 ]. The second study, a health economic evaluation of the E-MOTIVE trial in four sub-Saharan African countries, reported their intervention was highly cost-effective, incurring only $118.10 per DALY averted [ 93 ]. CHEC-E scores and certainty of effectiveness data of the 4 bundle evaluations varied (Table D in S7 Appendix and Table V in S8 Appendix ).
Care bundles are a complex strategy where multiple interventions are used simultaneously [ 90 ]. Three evaluations using effectiveness trial data [ 91 – 93 ], and 1 modelling study [ 94 ] considered care bundles for PPH-related care ( Table 5 ). Two evaluated a combination of preventative and treatment interventions [ 92 , 94 ]. A study in Niger reported a bundle of preventative uterotonics, improvised blood loss measurement and tiered treatment responses was highly effective and incurred low incremental costs [ 92 ]. A USA study showed that PPH preparedness and treatment strategies could be bundled, improving maternal health outcomes and reducing costs [ 94 ].
One high-quality modelling study assessed the cost-effectiveness of an emergency interfacility referral and transfer service for women and neonates in rural Madagascar [ 89 ]. The service, which assessed and transported patients from primary level care to secondary level facilities was estimated to be highly cost-effective, incurring $19.82 per additional life year saved for women with PPH. However, these results should be interpreted cautiously as the sub-analysis of women with PPH included only 46 patients, and survival rates were based on expert opinion rather than primary data.
Two modelling studies assessing cell salvage during cesarean section were identified [ 83 , 84 ]. In the UK study, cell salvage in cesarean section compared to no cell salvage incurred $13,714 per donor blood transfusion avoided, and was therefore unlikely to be cost-effective [ 83 ]. In the USA study, a broader study perspective, longer time horizon, and higher effectiveness estimates were used [ 84 ]. The authors concluded that cell salvage when women were at high risk of haemorrhage was likely cost-effective at their stated thresholds [ 84 ]. Both studies were high quality, though possible bias related to treatment effects could not be assessed due to the considerable variability of published effectiveness estimates [ 86 , 87 ].
One trial-based study in the Netherlands assessed whether women with acute anaemia following PPH should be transfused to a target haemoglobin or treated conservatively with iron and folic acid supplementation [ 85 ]. Although no clear cost-effectiveness threshold was set, the large costs required to lift women’s fatigue and quality of life scores by marginal amounts were not justified from the hospital perspective [ 85 ].
A high-quality modelling study in the USA attempted to identify the most cost-effective strategy for predelivery blood type, screening, and cross-matching [ 81 ]. Although some strategies were successful at reducing emergency-release transfusions, the upfront costs of predelivery testing outweighed the benefits; the best strategy was no routine admission testing [ 81 ]. Possible bias related to treatment effects could not be assessed (Table U in S8 Appendix ).
Five studies assessed various strategies for optimising blood product management in women with PPH [ 81 – 85 ]. The first, an economic evaluation from trial data, assessed blood resuscitation for severe PPH guided by point-of-care viscoelastic testing (PCVT) compared to empiric blood product resuscitation in the USA [ 82 ]. The intervention group experienced lower volumes of estimated blood loss, lower hysterectomy and intensive care unit admission rates, and significantly lower hospital costs [ 82 ]. However, the study was low quality, may be at risk of bias related to treatment effect (Table U in S8 Appendix ) and it is unclear if the device-associated costs were included in the analysis.
Two modelling studies assessed the cost-effectiveness of treating PPH with UBT devices in Kenya and India [ 78 , 79 ]. The study in Kenya assessed the “Every Second Matters” (ESM) UBT against standard care either with or without any form of uterine packing [ 79 ]. In this setting, utilising ESM-UBT was considered cost-effective compared to standard care, regardless of whether uterine packing was able to be performed. Incremental costs were between $30.26 and $231.64 per DALY averted, depending on the unit cost and which comparator was chosen [ 79 ]. This model was moderate quality and may be at risk of bias related to treatment effects (Table U in S8 Appendix ). The Indian study compared the relative cost-effectiveness between 3 UBT models—the ESM-UBT, the Bakri-UBT, or an improvised condom-UBT [ 78 ]. Although the base case of this analysis estimated the usage of the ESM-UBT to be the dominant strategy, the probabilistic sensitivity analysis (PSA) showed large uncertainty, and the authors could not conclude that switching from improvised devices to ESM-UBT in India would be cost-effective [ 78 ]. The assessment of one further device was identified, the novel Butterfly device for uterine compression [ 80 ]. This device is not yet utilised outside of trials and the results are reported in Table 3 .
One evaluation of an effectiveness study [ 76 ] and 1 modelling study [ 77 ] assessed non-pneumatic anti-shock garment (NASG) for women with obstetric haemorrhage in LMICs. One study in Zimbabwe and Zambia compared the cost-effectiveness of applying NASGs to patients in primary health centres awaiting definitive management at a referral centre, to waiting until patients arrived at the referral centre to apply NASG [ 76 ]. Results differed between sites, but on average the costs of distributing and training users of NASG were largely offset by the reduction in other health resources utilised when delaying application. A small incremental cost of $27.64 was incurred for each DALY averted by the intervention [ 76 ]. The second study in Egypt and Nigeria assessed the use of NASG within tertiary hospitals, against routine care without the temporising device [ 77 ]. In Egypt, NASG improved health outcomes and reduced costs, while in Nigeria it improved health outcomes at a modest incremental cost of $3.97 per DALY averted [ 77 ]. These evaluations were moderate [ 76 ] to high [ 77 ] quality. Bias related to treatment effects could not be assessed (Table U in S8 Appendix ).
Four modelling evaluations assessed tranexamic acid for PPH treatment—2 in the USA [ 72 , 73 ] and 2 in LMICs (India [ 74 ], Nigeria, and Pakistan [ 75 ]). Both USA studies considered women with PPH following either vaginal birth or cesarean section, concluding that tranexamic acid would likely avert laparotomies, deaths, and reduce healthcare and societal costs, whether on a short-term time horizon [ 73 ] or a lifetime horizon [ 72 ]. The LMIC-based studies were similarly positive for health outcomes but estimated modest incremental costs. These ranged from $86.03 per QALY gained in India [ 74 ] to $239.49 in Nigeria [ 75 ]. All 4 evaluations were high quality, and the effectiveness estimates used were consistent with systematic reviews.
Two studies assessed PPH treatment with misoprostol in settings where other treatment options are limited ( Table 4 ) [ 57 , 71 ]. The first analysed the cost-effectiveness of training traditional birth attendants (TBAs) to recognise and treat PPH with per-rectal misoprostol in sub-Saharan Africa [ 71 ]. They found it could prevent 1,647 cases from progressing to severe PPH, simultaneously saving $160,922 per 10,000 births [ 71 ]. The other study in India assessed a similar strategy of unskilled birth attendants diagnosing and treating PPH with misoprostol [ 57 ]. In this case, the health benefits (not as large as preventative strategies but preferable to no intervention) incurred a low incremental cost of $7.70 per DALY averted [ 57 ]. The studies were moderate [ 57 ] to high [ 71 ] quality. Bias related to treatment effects could not be assessed (Table U in S8 Appendix ).
One study in a USA hospital examined the effects of introducing the Triton system—combining colorimetric and gravimetric analysis for quantitative assessment of postpartum blood loss at all births—to improve PPH diagnosis ( Table 3 ) [ 70 ]. Triton implementation increased the proportion of women diagnosed with PPH. They calculated it would result in savings, though the analysis focussed on few types of cost.
The remaining 3 evaluations of preventative interventions related to surgical methods for minimising PPH [ 67 – 69 ]. The effectiveness of these interventions is debated. In addition, all 3 scored low on CHEC-E and pertained to a small proportion of women at risk of PPH (those with placenta accreta spectrum and those at high risk of PPH undergoing cesarean section). We therefore summarised their results separately in S12 Appendix .
One study assessed the cost-effectiveness of using a negative intrauterine pressure suction device alongside active management of the third stage of labour to minimise blood loss and prevent PPH in women following low risk vaginal births in India [ 66 ]. The study reported that using the device resulted in reduced blood loss, reduced rates of PPH (0.49% versus 1.81%, p < 0.001) and reduced hospital expenditure on blood products. Although this study met inclusion criteria for this review, the health economic analysis was extremely limited and only costed blood product usage. In addition, the underlying effectiveness evidence for negative intrauterine pressure devices remains uncertain and large-scale multicentre clinical trials are required [ 66 ].
A health technology assessment completed in UK hospitals evaluated the use of glyceryl trinitrate for the management of retained placenta to reduce the requirement for invasive procedures and the risk of PPH [ 64 , 65 ]. It was not effective at reducing the rate of manual placenta removal, or subsequent PPH, and was therefore not cost-effective [ 64 , 65 ].
Three high-quality studies assessed the impact of tranexamic acid for PPH prevention in high-income settings, 1 modelling study in women (vaginal birth or cesarean section) in the United States [ 59 ], and 2 economic evaluations of effectiveness studies in either women undergoing vaginal birth [ 60 ] or cesarean section [ 61 ] in France. In all 3 studies, the tranexamic acid strategy was either deemed cost-effective [ 61 ] or dominant [ 59 , 60 ]; however, there were some differences in the results. In the United States-based model, the impact on cost savings and avoided morbidity were substantial, while in the French studies, there were little differences in the overall effectiveness and costs between tranexamic acid and standard care strategies. The 3 studies scored high on CHEC-E, and 2 were based on effectiveness estimates either consistent [ 59 ] with systematic review estimates [ 7 , 62 , 63 ], or may have underestimated the effectiveness of tranexamic acid [ 60 ], thereby providing conservative cost-effectiveness estimates. The effectiveness estimates in one study were unable to be assessed, as they used a composite effectiveness measure not reported in previous systematic reviews [ 61 ].
In other settings, a Ugandan study estimated advanced misoprostol distribution would improve health outcomes at small incremental costs from the governmental perspective ($98.80 to $537.91 per DALY averted) or dominate under a societal perspective [ 54 ]. An analysis across sub-Saharan Africa similarly concluded community misoprostol use would incur additional costs and reduce maternal deaths [ 55 ]; they described the strategy as cost-effective without specifying a threshold. The final study assessed community or hospital misoprostol usage in international settings where injectable uterotonics are unavailable [ 53 ]. They concluded that either strategy of utilising misoprostol in the community alone, or in both hospitals and community, would be dominant [ 53 ]. As with the previous misoprostol studies, the underlying effectiveness estimates of all 3 studies was inconsistent with systematic reviews (Table O in S8 Appendix ) [ 30 , 37 ].
Six modelling studies analysed cost-effectiveness of misoprostol [ 52 – 57 ]. An Indian study assessed preventative misoprostol administered by village health workers (VHW) during home births compared to no uterotonics. They reported this strategy could reduce maternal deaths by 38% and cost $1,812.44 per death averted [ 56 ]. Another comparing misoprostol as prevention or treatment to no uterotonics found preventative misoprostol was the most effective strategy, incurring $218.26 per disability-adjusted life year (DALY) averted compared with misoprostol as treatment [ 57 ]. A third Indian study considered adding community-based distribution of misoprostol to other service and infrastructure upgrades [ 52 ]. The estimated maternal mortality reduction compared to other upgrades alone was modest (6.9% to 12.3%), but estimated to be cost-saving over a lifetime horizon [ 52 ]. These 3 studies assumed misoprostol to be significantly more effective than indicated in systematic reviews (Table O in S8 Appendix ) and may therefore overestimate its cost-effectiveness [ 30 , 37 ].
Five studies assessed oxytocin compared to no uterotonics, non-injectable uterotonics, or other formulations of oxytocin [ 47 – 51 ]. Heterogeneity in their settings, quality and completeness of methods, prevented meaningful comparisons between studies. One of the studies was on an inhalable form of oxytocin still under investigation, and the other 4 studies were each based on effectiveness estimates inconsistent with more recent evidence. For completeness, the results of each of these studies are summarised in S11 Appendix ; however, the most complete assessment of the cost-effectiveness of oxytocin appear as part of the UK health technology assessment from 2019 discussed earlier [ 30 , 31 ].
Seven studies, including 3 evaluations from effectiveness studies [ 39 – 41 ] and 4 models [ 42 – 45 ], assessed carbetocin against oxytocin for cesarean section only. Five concluded that carbetocin was the favourable option being either dominant, in the UK [ 41 , 44 ] and Australia [ 40 ], or cost-effective in Peru [ 42 ] and Ecuador [ 43 ]. Two of these studies (including 1 assessed as low quality) [ 41 ] may have over-estimated the relative effectiveness of carbetocin [ 41 , 42 ] (Table O in S8 Appendix ). A Malaysian study concluded carbetocin was more effective but more costly for cesarean section [ 45 ] and did not state a cost-effectiveness threshold to interpret this finding. In contrast, one study concluded that oxytocin was the favourable option for the UK [ 39 ], although this study was low quality and used effectiveness data that may underestimate the effectiveness of carbetocin [ 39 ] (Table O in S8 Appendix ). For vaginal birth, one UK study compared carbetocin and oxytocin in hospitals, concluding carbetocin was likely to be dominant [ 46 ]. The study was high quality, and the effectiveness estimates used in the model were consistent with recent systematic reviews.
Five studies used decision analytic models to compare carbetocin with oxytocin for both vaginal birth and cesarean section [ 32 – 36 ]. Models set in Canadian [ 35 ], and Chinese hospitals [ 34 ], and Indian primary, secondary, and tertiary health facilities [ 32 ] found that implementing carbetocin was dominant (both cost-saving and more efficacious) compared with oxytocin or misoprostol. A Colombian study concluded carbetocin would be either cost saving or cost-effective for cesarean section but not for vaginal birth [ 33 ]. A study in the Philippines found carbetocin was not cost-effective for either mode of birth [ 36 ]. However, three of these models [ 33 , 35 , 36 ] used effect estimates that may have either under- [ 33 , 36 ] or over-estimated [ 33 , 35 ] the relative effectiveness of carbetocin [ 30 , 37 , 38 ] (Table O in S8 Appendix ). The 2 studies using effect estimates consistent with current evidence both found carbetocin to be dominant [ 32 , 34 ].
Twenty-six (26/56) studies focused on preventative uterotonics ( Table 2 ). A high-quality health technology assessment considered cost-effectiveness of uterotonics for PPH prevention in United Kingdom (UK) hospitals in 2019 [ 30 , 31 ]. For vaginal birth, carbetocin was the most effective—incurring an incremental $1,530.93 per extra PPH > 500 ml averted when switching from oxytocin. For cesarean section, carbetocin was the second most effective and least costly, while misoprostol plus oxytocin was the most effective. This meant that switching from carbetocin to misoprostol plus oxytocin for cesarean section would incur an additional $4,091.60 per case of PPH ≥ 500 ml avoided. Mixed findings across analyses and a degree of uncertainty in the results led the authors to recommend against changing current practice (oxytocin for all births).
The top half of this figure represents the timing and order of commonly utilised preventative, diagnostic, and treatment interventions for postpartum haemorrhage. The bottom half of this figure displays the types of interventions assessed in the 56 studies we identified. *Not a WHO recommended intervention. # One study is listed twice as it included assessments of both uterotonics for prevention and uterotonics for PPH treatment. Number of studies identified per intervention type is shown in parenthesise. GTN, glyceryl trinitrate; IV, intravenous; NASG, non-pneumatic anti-shock garment; NIPD, negative intrauterine pressure device; PPH, postpartum haemorrhage; WHO, World Health Organization.
Searches identified 3,993 citations, of which, 56 were eligible for inclusion ( Fig 1 ; see Table A in S5 Appendix for excluded studies). Two additional citations were identified after reference review. Of the 58 total eligible citations, 2 (2/58) were found to be additional publications from the same studies meaning only 56 unique studies were identified. Of the included studies, 33 (33/56) assessed PPH prevention interventions, 1 (1/56) assessed prevention versus treatment, 1 (1/56) assessed a diagnostic method alone, 17 (17/56) assessed PPH treatments, and 4 (4/56) assessed combinations (bundles) of prevention and treatment, or diagnosis and treatment ( Fig 2 ). Studies were published between 2006 and 2024 and were conducted in low-income (3/56), lower-middle income (14/56), upper-middle income (8/56), and high-income (24/56) countries. Seven (7/56) studies were completed across multiple income level settings. Thirty-four studies (34/56) were set in hospitals, 13 (13/56) studies assessed interventions across multiple settings, 3 (3/56) studies assessed interventions for home births, 2 (2/56) studies assessed an intervention at a primary health setting, and 4 (4/56) studies did not clearly define which health facilities the studies were set. Of the 56 economic evaluations, 38 (38/56) were models, and 18 (18/56) were part of effectiveness trials. Study characteristics are summarised in Table 1 , including the perspective, time horizon, and overall CHEC-E score for each study. Most studies were conducted and reported comprehensively, with 33 (33/56) studies assessed as high quality on CHEC-E, 12 (12/56) moderate, and 11 (11/56) low (detailed study characteristics are presented in Table A in S6 Appendix and full CHEC-E assessments in Tables A–D in S7 Appendix ). S8 Appendix provides comparisons of effectiveness data used in each economic analysis, against estimates from published effectiveness reviews. Economic studies on interventions that are recommended by WHO (29 recommendations) were mapped to each recommendation by study country and country income level ( S9 Appendix ). For 16 (16/29) WHO recommended interventions, we identified zero economic evaluations. For 1 (1/29) recommendation, we identified studies from high-income countries only, and for 12 (12/29) recommendations, we identified corresponding studies from a range of high-, middle-, or low-income countries.
Discussion
This is, to our knowledge, the first systematic review examining the cost-effectiveness of interventions for PPH across the continuum of prevention, diagnosis, treatment, or combinations of these. We identified 56 studies—approximately half (24 studies) were conducted in high-income settings. Despite this considerable body of economic evidence for PPH-related care, the interventions, evaluation methodologies, time horizons, and perspectives varied considerably between studies. Acknowledging this heterogeneity, some patterns emerged. Currently, no injectable uterotonic agent or combination is universally dominant from a cost-effectiveness perspective. Across 15 studies assessing carbetocin against oxytocin, studies were split as to whether carbetocin improved health outcomes while incurring further costs or saved costs. The single UK study that considered the full range of available uterotonics did not recommend switching away from oxytocin [30,31].
Six studies found that when injectable uterotonics are unavailable using misoprostol is consistently cost-effective over no uterotonic [52–57]. However, the quality of studies was variable, and effectiveness estimates underpinning the cost-effectiveness calculations are inconsistent with recent estimates [30,37]. Studies of tranexamic acid for either prevention or treatment scored high on CHEC-E, used appropriate effectiveness estimates and concluded that the addition of tranexamic acid either dominated or was cost-effective across a range of settings [59–61,72–75], acknowledging that in some circumstances, the magnitude of the health benefit was small [61]. The 4 studies assessing the combination of multiple interventions into care bundles—variably involving prevention, diagnosis, and treatment components—suggest this can provide good economic value [91–94], particularly in LMICs [92,93].
Reviewing the effectiveness data used in each study’s cost-effectiveness calculations provided valuable insights. In some economic evaluations, the magnitude of an intervention’s effect was dramatically different to effect estimates reported in reputable systematic reviews. For example, one study’s base-case evaluation was centred around 600 μg of oral misoprostol being more than 6 times more effective at preventing transfer to hospital due to PPH, than 10 IU (international units) of intramuscular oxytocin for women in maternity huts in Senegal [47]. The trial this estimate was based on only had a single event for this outcome [95]. In other studies, assessing the cost-effectiveness of misoprostol administered by skilled staff, misoprostol was repeatedly assumed to lower the risk of PPH by 50% to 80% compared to no medication [52,56,57]. More recent effect estimates from systematic reviews indicate that it likely reduces the risk of PPH by approximately 25% [37]. The difference in assumed effectiveness can greatly change the number of expected outcomes, and therefore how cost-effective the intervention is perceived to be. For illustration, in a cohort of 10,000 women giving birth, with a baseline rate of PPH of 6%, altering the assumed effectiveness of misoprostol from reducing risk of PPH by 50% down to 25% would result in an extra 150 cases of PPH, and considerably change the costs used in the evaluation.
Previous systematic reviews have focused only on the cost-effectiveness of uterotonics for prevention [11], intrauterine devices for treatment [12], or tranexamic acid for treatment [13]. Similar to the 2019 review of preventative uterotonics, we were unable to clearly identify a particular agent that performed better from an economic perspective, despite a doubling of included studies. In comparison with the 2020 review on uterine tamponade devices, we identified one further economic evaluation [78], though economic data on this intervention remains sparse. Since the 2021 review on tranexamic acid for PPH treatment, one further economic evaluation has been published [74], and its results are compatible with the previous conclusion—that it is likely either cost-saving or cost-effective.
Although there are robust, comparative effectiveness data for the use of injectable uterotonics in PPH prevention [37], there remains no universally dominant choice from an economic perspective. The most economical choice of uterotonic is context-specific—considering trade-offs in effectiveness, side-effect profile, cost, availability, heat-stability, and the presence of trained providers to administer them. By comparison, the economic rationale for incorporating tranexamic acid into PPH treatment is more straightforward. It is recommended by WHO [23,29] with compelling evidence of benefits [96], and it has been demonstrated to be cost-saving or cost-effective in diverse settings [72–75]. In contrast, tranexamic acid for PPH prevention is not currently recommended by WHO. Effectiveness evidence of its prophylactic use suggests some benefit in preventing postpartum blood loss [7,62,63,97], noting that the 3 economic evaluations suggesting that it is a dominant or cost-effective strategy were from high-income settings [59–61]. While the addition of single agents to PPH prevention or treatment protocols can offer clinical and economic benefits, the implementation of combining therapies using a care bundle approach offers another path forward. There is high-quality economic evidence that care bundles can reduce the burden of PPH in LMIC [90,98] at modest incremental costs [93].
This review identified large gaps in the health economic literature that require further research. Several pivotal PPH interventions—such as oxytocin for treating PPH and quantitative measurement of postpartum blood loss—do not have high-quality economic data. Other interventions, such as PPH prevention using misoprostol have numerous economic evaluations, yet all were based on effectiveness estimates that have been superseded and are therefore of limited utility. Some interventions such as UBT and NASG demonstrated good value economically but have only been evaluated in a small number of studies and settings.
While effectiveness evidence forms the basis for recommendations, cost-effectiveness and resource use data play a key role in their implementability, especially so for resource-limited settings. Thus, effectiveness trials on PPH prevention, diagnosis, and management should have economic evaluations routinely embedded, helping maximise their translation and impact. The E-MOTIVE trial [98], and its embedded economic evaluation [93,99] is a recent successful example. First-order approximations of how results could translate across settings are possible to an extent with modelling that accounts for different demographics, health system characteristics, baseline intervention coverage and costs. These can be used to inform local, context-specific policies, but should not be used in the absence of considering local acceptability and feasibility. Translatability of results from modelling and our ability to make evidence informed decisions based on their results would also benefit from the creation of more robust methods for evaluating the certainty of economic evaluations [100].
Strengths of this review include the broad search conducted on multiple databases, with no limitations on language or date, and the systematic methodology employed in screening studies, extracting data, appraising quality, and resolving disputes. In addition, only peer-reviewed, economic analyses including the assessment of both health and financial outcomes were included in this review. Beyond appraising study quality, we also compared the effectiveness data used in each study to up-to-date systematic reviews, to minimise the possibility that economic findings based on outdated effectiveness data had influenced our conclusions.
This review has several limitations. Firstly, it was completed without nominating an ideal setting. Without an ideal setting, no transferability assessment could be completed. It is therefore important that users of the collated and summarised evidence in this review consider the extent to which each study finding is applicable to their own (or a nominated) setting. Secondly, while some major sources of study uncertainty were interrogated through the CHEC-E assessment, the description of differences between study settings and methods, and the in-depth assessment of effectiveness parameters used, other possible sources of uncertainty were not able to be interrogated to the same extent. In particular, a summary of the costing parameters used, and model validation methods undertaken by each study were not completed beyond the requirements of the CHEC-E assessment tool. Thirdly, our review of effectiveness estimates relied on recent systematic reviews; more recent data from single randomised trials, not yet incorporated into meta-analyses, are thus not captured. Fourthly, by choosing to summarise economic evaluations across prevention, diagnosis and treatment of PPH this review has prioritised breadth over the depth of analysis in the main manuscript. More detail, such as an analysis of the most appropriate study perspective and time horizons would have been possible through single intervention reviews on a smaller group of studies. It is important that the results summarised in this review are interpreted in the context of the more detailed information contained in the supplementary materials, namely the study characteristics (S6 Appendix), CHEC-E scores (S7 Appendix), and the analysis of the effectiveness data used in each study (S8 Appendix). Fifthly, several included studies were relatively old—7 were published prior to 2014. While historical cost estimates can be adjusted to present value, the treatment choices, evidence to inform key model parameters, and methodologies used to conduct and report economic evaluations have evolved over time.
[END]
---
[1] Url:
https://journals.plos.org/plosmedicine/article?id=10.1371/journal.pmed.1004461
Published and (C) by PLOS One
Content appears here under this condition or license: Creative Commons - Attribution BY 4.0.
via Magical.Fish Gopher News Feeds:
gopher://magical.fish/1/feeds/news/plosone/
