A systematic review of the cost‐effectiveness of uterotonic agents for the prevention of postpartum hemorrhage

Several uterotonic options exist for prevention of postpartum hemorrhage (PPH); hence, cost‐effectiveness is an important decision‐making criterion affecting uterotonic choice.

T A B L E 1 Description, pharmacokinetics, and illustrative relative unit costs of uterotonic agents. a

| Data extraction and quality assessment
Cochrane methodology for economic evaluations was used. 11 Two researchers (TL and ER) independently assessed for inclusion all potentially eligible studies. Any disagreements were resolved through discussion or by consulting a third author. A data extraction form was adapted from the guidance for NHS EED abstracts of economic evaluations. 12 For each included study, two authors independently extracted information on type of cost evaluation (study design), sources of effectiveness data, type of costs and their source, sources of outcome valuations, and type of adopted analytical approach. Where available, Incremental Cost-Effectiveness Ratios (ICERs) were extracted (an ICER is the ratio of the difference in costs between an intervention or treatment and a specified comparator to the difference in effectiveness or outcome between that intervention and the specified comparator).
Quality of identified studies was systematically evaluated using the Consensus Health Economic Criteria (CHEC) checklist. 13 Assessments were compared and disagreements resolved through discussion or consulting a third author.

| Data synthesis
We developed a conceptual framework on the possible cost consequences associated with use of a uterotonic agent to prevent PPH ( Fig. 1).
For identified studies, we tabulated the characteristics and findings of individual economic evaluations and supplemented it with a brief narrative summary of the findings. The currency and price year applicable to measures of costs in each original study were reported alongside measures of costs, incremental costs, and incremental cost-effectiveness.
Findings were summarized according to mode of birth (vaginal birth or cesarean delivery) and birth setting (hospital or community).

| Characteristics of included studies
The combined deduplicated search yield was 180 records. We additionally identified two records (one conference abstract and one commissioned report) through searching the reference lists of included papers. Out of 182 retrieved citations, 168 studies were assessed as irrelevant on title and abstract screening and nine full text retrievals contained no cost-effectiveness data, leaving 15 studies that met the eligibility criteria of this review (Fig. 2). One unpublished report identified in the reference list of an included study could not be retrieved: two were conference abstracts and contained limited data. 14,15 Included studies were conducted in Columbia, 15 Ecuador, 16 India, 17,18 Malaysia, 19 Mexico. 14 Peru, 20 Senegal, 21 Tanzania, 22 Uganda, 23 and the UK. [24][25][26][27] One of the studies was an international study with a hypothetical cohort. 28 Eleven of these studies were conducted from 2011 to 2018 and four studies were conducted from 2007 to 2010. 14 23,25,26 in two studies (abstracts) it was unclear whether they were funded or not, 14,15 and one study did not disclose funding sources. 19 The relevant healthcare perspective was used in all studies except in one study, which adopted a WHO perspective. A majority of the studies were model-based, using decision analytical models (decision trees), with the exception of two studies that used data directly obtained from randomized controlled trials, 14,21 one observational study, 25 and a service composite clinical and financial analysis study. 27 Various measures (condition-specific and generic) were used to measure health outcome of uterotonics including incidence and cases of PPH, use of additional uterotonics, mortality, probability of mortality, referral to higher-level health facility, adverse effects, quality-adjusted life years (QALYs), and disability-adjusted life years (DALYs). One study was a cost-effectiveness analysis of several uterotonics based on a network meta-analysis (NMA) adopting a UK perspective. 24 Six studies evaluated the cost-effectiveness of misoprostol versus third stage management without any uterotonics (five studies) or oxytocin (one study), all of which were conducted in settings with low access to facility births. 17 per additional PPH greater than or equal to 1000 mL avoided. These findings were similar for hospital and community settings in the UK.
Misoprostol was found to be the costliest uterotonic agent in this setting when adverse events were considered.

| Which uterotonic agents are cost-effective for preventing PPH at cesarean delivery?
Eight cost-effectiveness analyses and one service evaluation study contributed data. 14 were both more cost-effective than misoprostol or oxytocin alone.
The eight other studies that evaluated economic outcomes associated with carbetocin at cesarean delivery all compared carbetocin F I G U R E 1 Diagram of possible consequences associated with use of a uterotonic agent. (100 μg) to oxytocin (5 or 10 IU, where dose was reported). Seven studies concluded that carbetocin was cost-effective at cesarean delivery compared with oxytocin. [14][15][16]19,20,25,26 However, in some of these studies 19,20,26 investigators reported that uncertainty around costs and other input data made it difficult to evaluate costeffectiveness with any certainty. We considered the quality of the evidence provided by these studies to be low, owing to a lack of sensitivity analyses and incomplete description of the methods of outcome and costs measurements in most studies. In addition, at least two of the studies from middle-income countries favoring carbetocin were funded by the manufacturer and, as such, we interpreted findings from these studies and those of the conference abstracts with limited data 14,15 cautiously owing to the potential risk of bias and imprecision.

Increasing costs
In the remaining study, a UK service evaluation study, 27

| Uterotonic agents in community settings without skilled birth attendants
The cost-effectiveness of misoprostol was evaluated in settings with low access to modern birth facilities (lack of skilled birth attendants, inadequate transport and storage facilities, or oxytocin not available) in six studies. 17,18,[21][22][23]28 These studies were of moderate to high quality according to the CHEC checklist, 13 and most used a model-based approach to estimate incremental costs of introducing misoprostol to prevent PPH in these settings.
Four studies evaluated misoprostol as a 600μg dose (oral or sublingual), 17,18,21,23 one used a 200μg dose, 28 and one used a 1000μg dose rectally administered. 22 In most studies administration of misoprostol was undertaken by lay health workers; in one study in Uganda it was distributed prenatally to pregnant women for selfadministration following birth. 23

| Uterotonic agents in hospital settings where oxytocin quality cannot be guaranteed
We found no research evidence that directly addresses this question.
However, in the UK NMA cost-effectiveness analysis, if the option of oxytocin alone or in combination with other agents was removed from the study analyses, findings suggest that carbetocin might be the most cost-effective uterotonic for vaginal birth (and cesarean delivery) in the UK context, followed by ergometrine (if adverse events were considered), or misoprostol (if adverse events were not considered).
These findings were uncertain and might not be generalizable to other settings, particularly settings where injectable uterotonics cannot be used.

| DISCUSSION
Cost-effectiveness studies on PPH prevention in different contexts were sparse given the range of uterotonic agents available. We found

F I G U R E 2 Flow diagram of search results and study selection.
T A B L E 2 Relative effects and resource implications of different uterotonic agents compared with oxytocin. a,b Oxytocin ( The cost of this resource has been estimated in one study as US $0.84 per birth in a low-resource setting. 16 T A B L E 2 (Continued) reasonable quality evidence that misoprostol compared with no uterotonic agent was cost-effective for preventing PPH at vaginal birth in settings with low access to modern birth facilities. With regard to preventing PPH at cesarean delivery, the evidence suggested that carbetocin might be more cost-effective than oxytocin, and that the next most cost-effective option after carbetocin might be the combination of misoprostol and oxytocin; however, these findings were fairly uncertain.
To our knowledge this is the only systematic review of the costeffectiveness of uterotonic agents for PPH prevention to date. We  This study is a systematic review of available formal economic analyses of uterotonics agents but is itself not an economic evaluation or analysis. Findings from our review highlight the lack of reliable evidence on the cost-effectiveness of these agents in different settings, and hence a well-conducted economic evaluation of all agents (particularly in low-resource settings) is needed. Such an analysis would serve as an important guide to decision and policy making in PPH prevention, particularly in LMICs.
Given the paucity of available evidence on uterotonic costeffectiveness, for the purposes of informing the WHO guideline update process, we created a logic model on cost-effectiveness considerations across the different uterotonic options to aid decision makers and other stakeholders ( Table 2). In this approach, we used up-to-date evidence on the relative effectiveness and safety of the different uterotonic agents relative to oxytocin from an updated Cochrane systematic review and NMA. 31 We tabulated the risk ratios of desirable and undesirable effects. In addition, we tabulated their potential cost consequences relative to oxytocin, including resource requirements related to staffing and training, equipment and infrastructure, staff time, supplies, and supervision and monitoring that are associated with these options ( Table 2). For uterotonic supply prices, we referenced costs from the British National Formulary, 32

CONFLICTS OF INTEREST
The authors have no conflicts of interest to declare.