Open Collections

UBC Faculty Research and Publications

Magnesium sulphate for fetal neuroprotection: a cost-effectiveness analysis Bickford, Celeste D; Magee, Laura A; Mitton, Craig; Kruse, Marie; Synnes, Anne R; Sawchuck, Diane; Basso, Melanie; Senikas, Vyta M; von Dadelszen, Peter Dec 19, 2013

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
52383-12913_2013_Article_2942.pdf [ 701.69kB ]
Metadata
JSON: 52383-1.0221432.json
JSON-LD: 52383-1.0221432-ld.json
RDF/XML (Pretty): 52383-1.0221432-rdf.xml
RDF/JSON: 52383-1.0221432-rdf.json
Turtle: 52383-1.0221432-turtle.txt
N-Triples: 52383-1.0221432-rdf-ntriples.txt
Original Record: 52383-1.0221432-source.json
Full Text
52383-1.0221432-fulltext.txt
Citation
52383-1.0221432.ris

Full Text

RESEARCH ARTICLE Open AccessMagnesium sulphate for fetal neuroprotection: acost-effectiveness analysisCeleste D Bickford1*, Laura A Magee2, Craig Mitton1, Marie Kruse3, Anne R Synnes4, Diane Sawchuck5,Melanie Basso6, Vyta M Senikas7, Peter von Dadelszen5 and on behalf of the MAG-CP Working GroupAbstractBackground: The aim of this study was to assess the cost-effectiveness of administering magnesium sulphate topatients in whom preterm birth at < 32+0 weeks gestation is either imminent or threatened for the purpose of fetalneuroprotection.Methods: Multiple decision tree models and probabilistic sensitivity analyses were used to compare theadministration of magnesium sulphate with the alternative of no treatment. Two separate cost perspectives wereutilized in this series of analyses: a health system and a societal perspective. In addition, two separate measures ofeffectiveness were utilized: cases of cerebral palsy (CP) averted and quality-adjusted life years (QALYs).Results: From a health system and a societal perspective, respectively, a savings of $2,242 and $112,602 is obtainedfor each QALY gained and a savings of $30,942 and $1,554,198 is obtained for each case of CP averted whenmagnesium sulphate is administered to patients in whom preterm birth is imminent. From a health systemperspective and a societal perspective, respectively, a cost of $2,083 is incurred and a savings of $108,277 isobtained for each QALY gained and a cost of $28,755 is incurred and a savings of $1,494,500 is obtained for eachcase of CP averted when magnesium sulphate is administered to patients in whom preterm birth is threatened.Conclusions: Administration of magnesium sulphate to patients in whom preterm birth is imminent is a dominant(i.e. cost-effective) strategy, no matter what cost perspective or measure of effectiveness is used. Administrationof magnesium sulphate to patients in whom preterm birth is threatened is a dominant strategy from a societalperspective and is very likely to be cost-effective from a health system perspective.Keywords: Magnesium sulphate, Fetal neuroprotection, Preterm birth, Cerebral palsy, Cost-effectivenessBackgroundCerebral palsy (CP) is associated with substantial health-care and societal costs, as well as a significant reduction inhealth-related quality of life for those with moderate tosevere levels of disability [1-3]. It is estimated that preterminfants constitute up to 32% of all cases of CP, with theoverall prevalence of CP being approximately 2 cases per1,000 births [4]. Antenatal administration of magnesiumsulphate for fetal neuroprotection has been shown to re-duce the incidence of CP among preterm infants [5-11].These findings have subsequently led to the publicationof clinical practice guidelines on the use of magnesiumsulphate for fetal neuroprotection in Canada and Australia,and endorsement of the use of magnesium sulphate forfetal neuroprotection by the American College of Obstetri-cians and Gynaecologists [12-14].The main purpose of this paper is to present the re-sults of a cost-effectiveness analysis on the use of mag-nesium sulphate for fetal neuroprotection. However, it isalso the first paper to report the per-patient cost to ad-minister magnesium sulphate antenatally specifically forthe purpose of fetal neuroprotection and the first topresent previously unpublished data on the lifetime costof CP stratified by level of physical disability.* Correspondence: celesteb@interchange.ubc.ca1School of Population and Public Health, Faculty of Medicine, University ofBritish Columbia, Vancouver, CanadaFull list of author information is available at the end of the article© 2013 Bickford et al.; licensee BioMed Central Ltd. This is an open access article distributed under the terms of the CreativeCommons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, andreproduction in any medium, provided the original work is properly cited.Bickford et al. BMC Health Services Research 2013, 13:527http://www.biomedcentral.com/1472-6963/13/527MethodsModel structureAdministration of magnesium sulphate was compared onlywith no treatment, as there are currently no other ante-natal therapies used to prevent CP among preterm infants.The decision tree shown in Figure 1 was used to model thecost-effectiveness of administering magnesium sulphate topatients in whom preterm birth at < 32+0 weeks gestationis imminent (i.e. certain to occur within 24 hours). A separ-ate decision tree (Figure 2) was constructed to model thecost-effectiveness of administering magnesium sulphate topatients in whom preterm birth at < 32+0 weeks gestationis threatened (i.e. could occur within 24 hours, but isnot certain to), as there is no infallible way to identify allpatients who will deliver within a given time period. Un-published data gathered through the Canadian PerinatalNetwork suggest that only 24.4% of patients who presentwith indications such as preterm labour, antepartumhemorrhage, or premature prelabour rupture of the mem-branes deliver within 24 hours of admission to hospital[15]. As such, clinicians are likely to err by “overusing”magnesium sulphate among patients who present withthreatened preterm birth. The decision tree shown inFigure 2 took into account: (i) that additional nursingcosts would be incurred for patients in whom pretermbirth at < 32+0 weeks gestation is threatened, as they wouldnot otherwise need the one-on-one care required for ad-ministration of magnesium sulphate; (ii) that some wouldultimately go on to deliver at more than 32+0 weeks gesta-tion; and (iii) that some would be eligible to receive a sec-ond course of treatment with magnesium sulphate if theyfailed to deliver after the first course of treatment. It wasassumed that re-treatment would only occur if there was ahigh degree of certainty that delivery was imminent andthat all patients who received a second course of treat-ment would deliver at < 32+0 weeks gestation.Maternal outcomes were excluded from the models, asthere are only a few for which statistically significant dif-ferences have been found between women who had mag-nesium sulphate administered antenatally and those whodid not [5-7]. These include transient hypotension, tachy-cardia, flushing, and nausea, which can all be reversed bydiscontinuing infusion of the drug. Note that the impacton the patient’s overall health utility is relatively small foreach of these outcomes and the time period spent in eachstate of reduced health utility is brief, thereby making anydifference in the quality-adjusted life years (QALYs) be-tween the two groups negligible. The calculation of thisdifference would therefore not contribute meaningfully tothis analysis.CP was the only neonatal outcome included in themodels, as it is the only outcome for which a statisticallysignificant difference has been found between neonateswho were antenatally exposed to magnesium sulphate forthe purpose of fetal neuroprotection and those who werenot [5-7]. CP has been stratified based on level of physicaldisability for the present analysis, with mild correspondingto levels I and II of the Gross Motor Function ClassificationSystem (GMFCS), moderate corresponding to level III, andsevere corresponding to levels IV and V [16].ProbabilitiesThe estimated probabilities for CP and each level of phys-ical disability are shown in Table 1. Data from four RCTswere pooled to obtain these probabilities (Table 2) [8-11].For the threatened preterm birth model (Figure 2), theprobability of delivery within 24 hours of initiation of a sin-gle treatment with magnesium sulphate was estimated toFigure 1 Decision tree for the imminent preterm birth models. The decision tree used to compare standard care (no treatment) withadministration of magnesium sulphate to patients in whom preterm birth at < 32+0 weeks gestation is imminent.Bickford et al. BMC Health Services Research 2013, 13:527 Page 2 of 12http://www.biomedcentral.com/1472-6963/13/527be 0.2442 and the probability of re-treatment (given no de-livery after an initial course of treatment) was estimated tobe 0.6041 based on previously unpublished data from theCanadian Perinatal Network [15].Life expectanciesLife expectancies at birth were estimated using bothpublished median survival times and expert opinion(Table 3) [17,18]. It has been assumed that those withmild CP have a life expectancy at birth equal to that ofthe general population in Canada [19].Health-related quality of lifeThe Health Utilities Index Mark 3 (HUI-3) is a validated,multi-attribute health classification system used to measurehealth-related quality of life [2,3]. HUI-3 utility scores aregenerated using scores for eight health attributes (vision,hearing, speech, ambulation, dexterity, emotion, cognition,and pain) [20]. Data from two studies were combined usingweighted averages to obtain HUI-3 utility scores for indi-viduals with CP (Table 4) [2,3]. Note that utility valueswere estimated using expert opinion for the age ranges forwhich no data were available. HUI-3 utility scores forindividuals in the general population are shown inTable 5 [21,22].Cost to administer magnesium sulphateAll costs are presented in 2011 Canadian dollars [$1CAN=$1US = £0.62GBP]. The average cost (per 10,000 patients)to administer magnesium sulphate to patients in whompreterm birth at < 32+0 weeks gestation is imminent was es-timated to be $57.08 per patient. Variation exists in dosingregimens, therefore this estimate assumed an adminis-tration protocol similar to those outlined in the CanadianTable 1 Estimated probability of CP among children bornat < 32+0 weeks gestation and estimated probability ofeach level of physical disability given a diagnosis of CPMgSO4 No MgSO4 p-valueCP 0.0524 0.0736 0.0144Mild (GMFCS level I/II) 0.6364 0.6034 0.0552Moderate (GMFCS level III) 0.2597 0.2069 0.0004Severe (GMFCS level IV/V) 0.1039 0.1897 0.0001Figure 2 Decision tree for the threatened preterm birth models. The decision tree used to compare standard care (no treatment) withadministration of magnesium sulphate to patients in whom preterm birth at < 32+0 weeks gestation is threatened.Bickford et al. BMC Health Services Research 2013, 13:527 Page 3 of 12http://www.biomedcentral.com/1472-6963/13/527and Australian clinical practice guidelines (i.e. a 4 gramintravenous loading dose followed by a 1 gram per hourmaintenance infusion for a maximum of 24 hours) andtook into consideration (1) the drug acquisition cost, (2)any laboratory costs incurred, and (3) the incremental costof treating any adverse maternal or neonatal events associ-ated with administering the drug [12,13]. No incrementalnursing costs were added for monitoring patients in whompreterm birth at < 32+0 weeks gestation is imminent, asthey would already be receiving one-on-one nursing carein a delivery suite.A per patient drug acquisition cost of $34.35 was ob-tained from the pharmacy at an academic tertiary medicalcenter in Vancouver, Canada. This included a pharmacycost of $22.14 for one 1 liter bag of pre-mixed magnesiumsulphate at a concentration of 40 mg/mL in normal saline(includes the drug and mixing costs), a cost of $1.37 eachfor three 1 liter bags of normal saline, a cost of $5.37 forone set of IV tubing, and a cost of $2.73 for one IV can-nula (M. Tofan, Dipl Pharm, written communication,November 2011). Pre-mixing of the magnesium sulphateinfusion was preferred to minimize the risk of dosage er-rors. It was assumed that each patient would already havean intravenous drip in place and that pumps would bereadily available on the delivery unit.Laboratory costs would only be incurred if there wereevidence of maternal toxicity. It is anticipated that a max-imum of 10% of patients would have one magnesiumsulphate serum level measurement during the administra-tion period, based on the proportion of women who dis-continued infusion of the drug in the Magpie trial and whowould have had serum levels measured if treated in a highincome country [23]. The estimated cost for this measure-ment is $19.73 and is comprised of a $7.00 blood collectionfee, a $5.00 handling fee, and a $7.73 laboratory fee (D.Sawchuck, PhD, written communication, August 2011).It was assumed that the unlikely event of maternal re-spiratory depression would occur in a maximum 0.6% ofpatients, based on data from the Magpie trial [23]. Itwas further assumed that all of these patients wouldneed to receive calcium gluconate and be intubated andventilated for 24 hours. The estimated cost for this treat-ment is $2,884.30 and is comprised of a cost of $4.30 fora one gram dose of calcium gluconate (in a ten millilitersingle use vial), a $280.00 airway management fee by ananesthesiologist, and $2,600.00 for 24 hours of criticalcare [24].Excessive exposure to magnesium sulphate may resultin neonatal hypotonia and respiratory depression [25,26].It was assumed that the unlikely event of neonatal respira-tory depression would occur in a maximum 0.6% of pa-tients, based on the need for bag-mask ventilation in asub-analysis of the BEAM trial [27]. This would requireintubation and the need for level III NICU care for aperiod of 24 hours. To account for the maximum possibleincremental cost, it was assumed that all neonates bornat < 32+0 weeks gestation would otherwise require level IINICU care and would not be intubated. The estimatedcost associated with this treatment is $576.03 and is com-prised of an incremental cost of $126.03 for physician bill-ing fees and an incremental cost of $450.00 for 24 hoursof critical care in a level III NICU [28].The average cost (per 10,000 patients) to administermagnesium sulphate to patients in whom preterm birthat < 32+0 weeks gestation is threatened was estimated toTable 3 Estimated life expectancies at birth for individuals with CP stratified by level of physical disability, as well asmaximum and minimum values used in the sensitivity analysesEstimated life expectancy (years) Min MaxMild (GMFCS level I/II) 81 71 91Moderate (GMFCS level III) 66 56 76Severe (GMFCS level IV/V) 25 20 30Table 2 Data used to calculate the probability of CPMgSO4 No MgSO4 SourceMarrett 352 336 8CP 22 30Mittendorf 30 29 9CP 3 0Crowther 533 513 10CP 36 42Mild (GMFCS level I/II) 21 21 aModerate (GMFCS level III) 12 15 aSevere (GMFCS level IV/V) 3 6 aRouse 1030 1107 11CP 41 74Mild (GMFCS level I/II) 28 49 bModerate (GMFCS level III) 8 9 bSevere (GMFCS level IV/V) 5 16 bTotal number of patients assessed and number with each outcome in each ofthe four trials. Note that only data on patients to whom magnesium sulphatewas administered for the specific purpose of fetal neuroprotection wasincluded in the present analyses.aAssumed with reference to the study protocol and detailed descriptions ofcharacteristics for each GMFCS level for children under the age of four.bD. Rouse, MD, written communication, March 2011.Bickford et al. BMC Health Services Research 2013, 13:527 Page 4 of 12http://www.biomedcentral.com/1472-6963/13/527be $897.08 per patient. It was assumed that patients inwhom preterm birth at < 32+0 weeks gestation is threatenedwould not already be receiving the one-on-one nursingcare required for administration of magnesium sulphate.Therefore, the estimated cost to administer magnesiumsulphate to this population included both the $57.08administration cost outlined above and $840.00 in nursingcosts to account for a 24 hour period of one-on-one nurs-ing care (D. Sawchuck, PhD, written communication,August 2011).Lifetime cost of cerebral palsyThe lifetime costs of CP (Table 6) were estimated usingthe life expectancies shown in Table 3 and the healthcare, productivity, and social annual attributable cost in-puts shown in Table 7. Detailed information on themethodology used to determine these annual attribut-able cost inputs has been outlined in a previous publica-tion [1]. The health care costs consisted of all primaryhealth care, hospital, and pharmaceutical costs. Theproductivity costs consisted of all costs associated withlost labour market productivity for adults with CP. Thesocial costs consisted of all costs associated with special-ized education, specialized housing, and lost labour mar-ket productivity for primary care providers of childrenwith CP. Health care and productivity costs were quanti-fied using patient-level data obtained from a Danishpopulation-based register, while social costs were esti-mated using a combination of register data, previouslypublished literature, and expert opinion.All costs were originally determined in 2006 DanishKroner (DKK). These values were converted to Canadiandollars (CAD) using the average Bank of Canada nominalexchange rate between January 1, 2002 and January 1,2011 (1 CAD= 4.9494 DKK). Inflation was then adjustedfor using Canadian Health and Personal Care ConsumerPrice Index (CPI) values for 2006 and 2011, which were105.9 and 117.1 respectively. The formula used for cur-rency conversion was: (DKK2006)*(1.0000 CAD)/(4.9494DKK) = (CAD2006). The formula used to adjust for infla-tion was: (CAD2006)*(CPI2011)/(CPI2006) = (CAD2011).AnalysisThe cost-effectiveness of antenatal administration of mag-nesium sulphate for fetal neuroprotection among pretermdeliveries at < 32+0 weeks gestation as compared to notreatment were analysed from both a health system per-spective and a societal perspective. Two measures of ef-fectiveness were explored in separate analyses: (1) cases ofCP averted and (2) QALYs. QALYs are a measure of effect-iveness that take into account both the length of life andthe level of health-related quality of life associated with aparticular outcome [29,30]. Lifetime costs were used in allanalyses.Incremental cost-effectiveness ratios (ICERs) werecalculated in the base-case analyses. ICERs give the add-itional cost per unit of effectiveness gained when onetreatment alternative is compared to another [29,30].Probabilistic sensitivity analyses were conducted usingTable 4 Mean HUI-3 utility scores for individuals with CP stratified by level of physical disabilityAge(years)Mild Moderate Severe Source(GMFCS level I/II) (GMFCS level III) (GMFCS level IV/V)Mean SE Mean SE Mean SE0-4 0.79 0.24 0.66 0.25 0.47 0.22 –5-12 0.72 0.26 0.54 0.27 0.24 0.22 –13-20 0.66 0.28 0.41 0.29 0.01 0.23 2,321-22 0.61 0.32 0.47 0.28 −0.01 0.22 –23-32 0.57 0.35 0.53 0.27 −0.04 0.21 3> 32 0.57 0.35 0.53 0.27 −0.04 0.21 –Table 5 Mean HUI-3 utility scores for the generalpopulationAge (years) General population SourceMean SE0-12 0.92 0.11 –13-15 0.90 0.15 2116-37 0.85 0.18 2138-54 0.83 0.01 2255-64 0.77 0.02 2265-74 0.80 0.01 2275-89 0.75 0.01 22Table 6 The undiscounted lifetime cost of CP stratified bylevel of physical disabilityHealth systemperspectiveSocietalperspectiveMild (GMFCS level I/II) 33,524 3,581,722Moderate (GMFCS level III) 70,923 4,500,094Severe (GMFCS level IV/V) 61,378 2,208,153Bickford et al. BMC Health Services Research 2013, 13:527 Page 5 of 12http://www.biomedcentral.com/1472-6963/13/527Monte Carlo simulation, producing incremental cost-effectiveness plots (ICEPs) and cost-effectiveness ac-ceptability curves (CEACs). ICEPs illustrate the ICERsgenerated when model inputs are simultaneously variedin a series of analyses (e.g. 10,000 simulations). CEACsillustrate the probability of a treatment option beingcost-effective across various willingness-to-pay values.Each willingness-to-pay value represents the maximumcost that a decision-maker is willing to pay for an add-itional unit of effectiveness [29,30].The results of the Monte Carlo analyses are based on10,000 simulations. Beta and Dirichlet distributions wereused for probabilities, gamma distributions were used forcosts, beta distributions were used for utility weights, andtriangular distributions were used for life expectancies[29]. The standard errors for all costs were estimated to be20% of their point estimates, to account for any uncer-tainty or regional variation in these values. The standarderrors for all utility weights were taken from the literature(Tables 4 and 5). Minimum and maximum values for lifeexpectancy estimates for individuals with CP were deter-mined using expert opinion (Table 3). The life expectancyfor the general population was not varied, as it was calcu-lated using population-based data. In accordance with theWorld Health Organization’s guide to cost-effectivenessanalyses, costs and QALYs were discounted at 3% annuallyfor the base case analyses, but further sensitivity analyseswere also conducted using a 6% annual discount rate forcosts and a 0% annual discount rate for QALYs [30,31].All analyses were carried out using TreeAge Pro 2012(TreeAge Software, Inc, Williamstown, MA).ResultsImminent preterm birthA treatment option that is less costly and more effectivethan its alternative is always considered to be cost-effectiveand is referred to as a dominant strategy [30]. The pointestimates listed in Tables 8 and 9 indicate that administra-tion of magnesium sulphate is a dominant strategy com-pared to the alternative of no treatment no matter whatmeasure of effectiveness or cost perspective is used. Theresults generated from a health system perspective indicatethat a savings of $2,242 is obtained for each QALY gainedand a savings of $30,942 is obtained for each case of CPaverted. The results generated from a societal perspectiveindicate that a savings of $112,602 is obtained for eachQALY gained and a savings of $1,554,198 is obtained foreach case of CP averted.The ICEPs generated through the probabilistic sensitiv-ity analysis illustrate that when all model inputs are varied,the majority of the computed ICERs fall into the quadrantof the incremental cost-effectiveness plane which indi-cates that administering magnesium sulphate is a dom-inant strategy compared to the alternative of no treatment(Figure 3). The CEACs indicate that the probability of ad-ministration of magnesium sulphate being cost-effective isover 99% for all willingness-to-pay values up to $100,000,no matter what measure of effectiveness or cost perspec-tive is used (Figure 4).Threatened preterm birthThe point estimates listed in Tables 10 and 11 indicatethat administration of magnesium sulphate is a domin-ant strategy compared to the alternative of no treatmentonly when a societal perspective is used. When a healthsystem perspective is used, it is both more effective andmore costly than the alternative of no treatment. The re-sults generated from a societal perspective indicate thata savings of $108,277 is obtained for each QALY gainedand a savings of $1,494,500 is obtained for each case ofCP averted. The results generated from a health systemperspective indicate that a cost of $2,083 is incurred forTable 7 The annual attributable health care, productivity, and social costs for children and adults with CP stratified bylevel of physical disabilityChildren ≤18 years Adults > 18 yearsHealth Productivity Social Total Health Productivity Social TotalMild (GMFCS level I/II) 1,279 0 51,146 52,425 146 16,991 23,905 41,042Moderate (GMFCS level III) 2,895 0 77,179 80,074 331 25,653 36,072 62,056Severe (GMFCS level IV/V) 3,100 0 87,278 90,378 355 28,992 40,792 70,139Table 8 ICERs calculated for the imminent preterm birth models using cases of CP averted as the measure ofeffectivenessAnalysis Method Cost Δ cost Cases w/o CP Δ cases w/o CP ICER StatusHealth system perspective MgSO4 $1,672 -$653 0.95 0.02 -$30,942 DominantNo MgSO4 $2,326 0.93Societal perspective MgSO4 $85,822 -$32,808 0.95 0.02 -$1,554,198 DominantNo MgSO4 $118,630 0.93Bickford et al. BMC Health Services Research 2013, 13:527 Page 6 of 12http://www.biomedcentral.com/1472-6963/13/527each QALY gained and a cost of $28,755 is incurred foreach case of CP averted.The ICEPs generated through the probabilistic sensitiv-ity analysis illustrate that when all model inputs are varied,the majority of the ICERs computed from a societal per-spective fall into the quadrant of the incremental cost-effectiveness plane which indicates that administration ofmagnesium sulphate is a dominant strategy compared tothe alternative of no treatment. Conversely, the majorityof the ICERs computed from a health system perspectivefall into the quadrant of the incremental cost-effectivenessplane which indicates that that administration of magne-sium sulphate is both more costly and more effective thanthe alternative of no treatment (Figure 5). The CEACsgenerated using a societal perspective indicate that theprobability of administration of magnesium sulphate beingcost-effective is over 99% for all willingness-to-pay valuesup to $100,000, no matter what measure of effectivenessis used. The CEACs generated using a health systemperspective indicate much lower probabilities of cost-effectiveness (Figure 6).DiscussionMain findingsAdministration of magnesium sulphate to patients inwhom preterm birth at < 32+0 weeks gestation is imminentis a dominant (i.e. cost-effective) strategy compared to thealternative of no treatment, no matter what measure ofTable 9 ICERs calculated for the imminent preterm birth models using QALYs as the measure of effectivenessAnalysis Method Cost Δ cost QALYs Δ QALYs ICER StatusHealth system perspective MgSO4 $1,672 -$653 26.6 0.3 -$2,242 DominantNo MgSO4 $2,326 26.3Societal perspective MgSO4 $85,822 -$32,808 26.6 0.3 -$112,602 DominantNo MgSO4 $118,630 26.3Societal Perspective, QALYsHealth System Perspective, Cases of CP Averted Health System Perspective, QALYsSocietal Perspective, Cases of CP AvertedFigure 3 ICEPs for the imminent preterm birth models. The black ellipses represent 95% confidence intervals.Bickford et al. BMC Health Services Research 2013, 13:527 Page 7 of 12http://www.biomedcentral.com/1472-6963/13/527effectiveness or cost perspective is used. Administration ofmagnesium sulphate to patients in whom preterm birthat < 32+0 weeks gestation is threatened is a dominant strat-egy compared to the alternative of no treatment from onlya societal perspective. When a health system perspective isused its cost-effectiveness is dependent on the decision-makers’ willingness-to-pay for an additional QALY or caseof CP averted. The National Institute for Health andClinical Excellence suggests that society’s willingness-to-pay threshold for new health technologies is approximately$30,000/QALY, making anything less costly a cost-effectiveoption from the viewpoint of society as a whole [32].Administration of magnesium sulphate to patients inwhom preterm birth at < 32+0 weeks gestation is threatenedhas a very high probability of being cost-effective at allwillingness-to-pay values over $10,000/QALY, no matterwhat cost perspective is used. Therefore, it is very likely tobe a cost-effective option from the viewpoint of society as awhole even though it is not a dominant strategy from allcost perspectives.Health System Perspective, Cases of CP AvertedSocietal Perspective, QALYsHealth System Perspective, QALYsSocietal Perspective, Cases of CP AvertedFigure 4 CEACs for the imminent preterm birth models. The blue line depicts the probability of magnesium sulphate being cost-effectiveacross all willingness-to-pay values up to $100,000. The red line depicts the probability of the alternative (no treatment) being cost effective acrossall willingness-to-pay values up to $100,000.Table 10 ICERs calculated for the threatened preterm birth models using cases of CP averted as the measure ofeffectivenessAnalysis Method Cost Δ cost Cases w/o CP Δ cases w/o CP ICER StatusHealth system perspective MgSO4 $2,055 $425 0.96 0.01 $28,755aNo MgSO4 $1,630 0.95Societal perspective MgSO4 $61,028 -$22,109 0.96 0.01 -$1,494,500 DominantNo MgSO4 $83,137 0.95aCost-effectiveness is dependent on the willingness-to-pay threshold of the decision-maker.Bickford et al. BMC Health Services Research 2013, 13:527 Page 8 of 12http://www.biomedcentral.com/1472-6963/13/527Strengths and limitationsThis is the first study in which the cost to administer mag-nesium sulphate specifically for fetal neuroprotection hasbeen estimated (previous studies, including the Cahillet al. cost-effectiveness analysis, have only reported/uti-lized the cost to administer magnesium sulphate for toco-lysis or preeclampsia prophylaxis). It is also the first studyto present detailed data on the lifetime costs of CP by levelof physical disability. The use of these data in the presentanalysis is important because the extent of care requiredfor individuals with different levels of disability varies con-siderably. It is also important because the lifetime cost ofCP is a key driver of the cost-effectiveness of administra-tion of magnesium sulphate.The analyses included models for both imminent andthreatened preterm birth because there is no infallibleway to predict which patients will deliver imminently.As such, clinicians are likely to err by “overusing” mag-nesium sulphate among patients who present with whatappears to be imminent preterm birth, but is in factthreatened preterm birth. The threatened preterm birthmodel included important information on additionalresources required to administer magnesium sulphateto those who would be likely to receive it in clinicalSocietal Perspective, QALYsSocietal Perspective, Cases of CP AvertedHealth System Perspective, Cases of CP Averted Health System Perspective, QALYsFigure 5 ICEPs for the threatened preterm birth models. The black ellipses represent 95% confidence intervals.Table 11 ICERs calculated for the threatened preterm birth models using QALYs as the measure of effectivenessAnalysis Method Cost Δ cost QALYs Δ QALYs ICER StatusHealth system perspective MgSO4 $2,055 $425 26.7 0.2 $2,083aNo MgSO4 $1,630 26.5Societal perspective MgSO4 $61,028 -$22,109 26.7 0.2 -$108,277 DominantNo MgSO4 $83,137 26.5aCost-effectiveness is dependent on the willingness-to-pay threshold of the decision-maker.Bickford et al. BMC Health Services Research 2013, 13:527 Page 9 of 12http://www.biomedcentral.com/1472-6963/13/527practice. However, note that the assumption that a sec-ond course of treatment with magnesium sulphatewould occur only in the case of imminent preterm birthmay not always reflect clinical practice.The estimation of model inputs is a common limita-tion of decision analytic models. For the present analysis,many of the utility weights for individuals with CP wereestimated due to a paucity of published information onthe health-related quality of life of this population. Un-certainty also remains around the probability inputs,given that they were obtained using aggregated data ra-ther than data from a single randomized controlled trial[33]. In this analysis the impact that such estimates haveon the precision of the model has been minimizedthrough the use of a robust probabilistic sensitivity ana-lysis, which allows the effect of uncertainty across all theparameters to be considered simultaneously.Another common limitation of decision analytic modelsis the uncertainty that surrounds the structure of themodel itself. For example, it should be noted that the clas-sification of CP as being mild, moderate, or severe requiresbroad generalizations and that this is only one of manyclassification systems that could have been utilized. Inaddition, perinatal death was excluded from the model, asthe authors feel that there is not enough evidence at thistime to indicate whether there is a difference in perinataldeath between the two groups or what the magnitude ofthis difference may be. However, it should be noted thatthe assumption of no effect on mortality has been recentlydisputed in a meta-analysis of observational studies, whichfound that the use of magnesium sulphate was associatedwith a decrease in both CP and perinatal mortality [34].With respect to these findings, if a randomized controlledtrial were ever to show that antenatal administration ofmagnesium sulphate decreased the probability of peri-natal death, then the cost-effectiveness of administeringmagnesium sulphate for fetal neuroprotection would onlybe improved.Health System Perspective, Cases of CP Averted Health System Perspective, QALYsSocietal Perspective, Cases of CP Averted Societal Perspective, QALYsFigure 6 CEACs for the threatened preterm birth models. The blue line depicts the probability of magnesium sulphate being cost-effectiveacross all willingness-to-pay values up to $100,000. The red line depicts the probability of the alternative (no treatment) being cost effective acrossall willingness-to-pay values up to $100,000.Bickford et al. BMC Health Services Research 2013, 13:527 Page 10 of 12http://www.biomedcentral.com/1472-6963/13/527Comparison to other studiesOne other cost-effectiveness analysis on the use of magne-sium sulphate for fetal neuroprotection has been identified[35]. The results obtained in this study are consistent withthe findings of our analysis in that administration of mag-nesium sulphate for fetal neuroprotection was found to beless costly and more effective than the alternative of notreatment from a societal perspective. However, our prob-abilistic sensitivity analysis produced much higher prob-abilities of magnesium sulphate being cost-effective. Anexplanation for this difference is that the cost of CP usedin the Cahill et al. study was likely underestimated. Theydid not take into account the severity of CP and their esti-mate of the lifetime cost of CP was taken from a singlestudy, whereas our analysis used previously unpublished,population-based data on the lifetime cost of CP stratifiedby level of physical disability. Furthermore, the life expect-ancies and utility values utilized in their analysis were likelyoverestimated. We have therefore built on the Cahill et al.study by inserting more precise estimates into our model,in addition to using several cost perspectives and measuresof effectiveness and investigating the cost-effectiveness ofimminent and threatened preterm birth separately.ConclusionsAdministration of magnesium sulphate to patients inwhom preterm birth is imminent is a dominant (i.e. cost-effective) strategy, no matter what measure of effectivenessor cost perspective is used. Administration of magnesiumsulphate to patients in whom preterm birth is threatenedis a dominant strategy from a societal perspective and isvery likely to be cost-effective from a health system per-spective. Clinicians and administrators should interpretthe findings of this study as providing reassurance that if adiagnosis of imminent preterm birth at < 32+0 weeks gesta-tion is made and magnesium sulphate is administered, butthe patient does not deliver, this “overuse” of magnesiumsulphate is still likely to have been cost-effective.Competing interestsThe authors declare that they have no competing interests.Authors’ contributionsCB contributed to the literature review, study design, and analysis. CB draftedthe first version of the manuscript and led the submission process. LMinitiated the collaborative project, as well as contributing to the study designand analysis and obtaining costing data. CM, MK, AS, DS, MB, VS, and PvDcontributed to the study design and analysis and obtained costing data. Allauthors contributed to the interpretation of the study findings and therevision and approval of the final manuscript.AcknowledgementsWe are grateful to the Danish Cerebral Palsy Register for providing us withinformation on the lifetime costs of CP, Steven Weiner and Dwight Rouse forproviding us with previously unpublished data from the BEAM trial, and LisaMasucci and Mohsen Sadatsafavi for their assistance with the modeldevelopment and analysis. The Canadian Institutes of Health Researchprovided funding for the MAG-CP (magnesium sulphate for fetalneuroprotection) project, of which this analysis is a component.The MAG-CP (Magnesium sulphate for fetal neuroprotection of the preterminfant) Working Group consists of DA DeSilva, LA Magee, and AR Synnes.Author details1School of Population and Public Health, Faculty of Medicine, University ofBritish Columbia, Vancouver, Canada. 2Department of Medicine, Division ofInternal Medicine, University of British Columbia, Vancouver, Canada. 3DanishInstitute for Health Services Research, Copenhagen, Denmark. 4Departmentof Pediatrics, Division of Neonatology, Children’s & Women’s Health Centreof British Columbia, Vancouver, Canada. 5Department of Obstetrics andGynecology, Division of Maternal Fetal Medicine, University of BritishColumbia, Vancouver, Canada. 6Children’s & Women’s Health Centre of BritishColumbia, Perinatal Health Program, Vancouver, Canada. 7Faculty ofMedicine, McGill University, Montreal, Canada.Received: 10 May 2013 Accepted: 5 December 2013Published: 19 December 2013References1. Kruse M, Michelsen SI, Flachs EM, Brønnum-Hansen H, Madsen M, Uldall P:Lifetime costs of cerebral palsy. Dev Med Child Neurol 2009, 51:622–628.2. Rosenbaum PL, Livingston MH, Palisano RJ, Galuppi BE, Russell DJ: Qualityof life and health-related quality of life of adolescents with cerebralpalsy. Dev Med Child Neurol 2007, 49:516–521.3. Young NL, Rochon TG, McCormick A, Law M, Wedge JH, Fehlings D: Thehealth and quality of life outcomes among youth and young adults withcerebral palsy. Arch Phys Med Rehabil 2010, 91:143–148.4. Drummond PM, Colver AF: Analysis by gestational age of cerebral palsyin singleton births in north-east England 1970–94. Paediatr PerinatEpidemiol 2002, 16:172–180.5. Conde-Agudelo A, Romero R: Antenatal magnesium sulphate for theprevention of cerebral palsy in preterm infants < 34 weeks’ gestation:a systematic review and metaanalysis. Am J Obstet Gynecol 2009,200:595–609.6. Costantine MM, Weiner SJ, Eunice Kennedy Shriver National Institute ofChild Health and Human Development Maternal-Fetal Medicine UnitsNetwork: Effects of antenatal exposure to magnesium sulphate onneuroprotection and mortality in preterm infants: a meta-analysis.Obstet Gynecol 2009, 114:354–364.7. Doyle LW, Crowther CA, Middleton P, Marret S: Magnesium sulphate forwomen at risk of preterm birth for neuroprotection of the fetus.Cochrane Database Syst Rev 2009, 3, CD004661.8. Marret S, Marpeau L, Zupan-Simunek V, Eurin D, Lévêque C, Hellot MF,Bénichou J, PREMAG trial group: Magnesium sulphate given beforevery-preterm birth to protect infant brain: the randomized controlledPREMAG trial. BJOG 2007, 114:310–318.9. Mittendorf R, Dambrosia J, Pryde PG, Lee KS, Gianopoulos JG, Besinger RE,Tomich PG: Association between the use of antenatal magnesiumsulphate in preterm labor and adverse health outcomes in infants.Am J Obstet Gynecol 2002, 186:1111–1118.10. Crowther CA, Hiller JE, Doyle LW, Haslam RR, Australasian Collaborative Trialof Magnesium Sulphate (ACTOMg SO4) Collaborative Group: Effect ofmagnesium sulphate given for neuroprotection before preterm birth: arandomized controlled trial. JAMA 2003, 290:2669–2676.11. Rouse DJ, Hirtz DG, Thom E, Varner MW, Spong CY, Mercer BM, Iams JD, WapnerRJ, Sorokin Y, Alexander JM, Harper M, Thorp JM Jr, Ramin SM, Malone FD,Carpenter M, Miodovnik M, Moawad A, O’Sullivan MJ, Peaceman AM, HankinsGD, Langer O, Caritis SN, Roberts JM, Eunice Kennedy Shriver NICHD Maternal-Fetal Medicine Units Network: A randomized, controlled trial of magnesiumsulphate for the prevention of cerebral palsy. N Engl J Med 2008, 359:895–905.12. Antenatal Magnesium Sulphate for Neuroprotection GuidelineDevelopment Panel: Antenatal magnesium sulphate prior to preterm birth forneuroprotection of the fetus, infant and child: national clinical practiceguidelines. Adelaide: The University of Adelaide; 2010.13. Magee LA, Sawchuck D, Synnes A, von Dadelszen P: SOGC Clinical PracticeGuideline. magnesium sulphate for fetal neuroprotection. J ObstetGynaecol Can 2011, 33:516–529.14. American College of Obstetricians and Gynecologists Committee onObstetric Practice, Society for Maternal-Fetal Medicine: Committee OpinionNo. 455: magnesium sulphate before anticipated preterm birth forneuroprotection. Obstet Gynecol 2010, 115:669–671.Bickford et al. BMC Health Services Research 2013, 13:527 Page 11 of 12http://www.biomedcentral.com/1472-6963/13/52715. Magee LA, von Dadelszen P, Allen VM, Ansermino JM, Audibert F, Barrett J,Brant R, Bujold E, Crane JM, Demianczuk N, Joseph KS, Lee SK, Piedboeuf B,Smith G, Synnes A, Walker M, Whittle W, Wood S, Lee T, Li J, Payne B, ListonRM, Canadian Perinatal Network Collaborative Group: The CanadianPerinatal Network: a national network focused on threatened pretermbirth at 22 to 28 weeks’ gestation. J Obstet Gynaecol Can 2011,33:111–120.16. Palisano R, Rosenbaum P, Bartlett D, Livingston M: Gross motor functionclassification system – expanded and revised. Hamilton: CanChild Centre forChildhood Disability Research; 2007.17. Hutton JL, Pharoah PO: Life expectancy in severe cerebral palsy. Arch DisChild 2006, 91:254–258.18. Hemming K, Hutton JL, Pharoah PO: Long-term survival for a cohort ofadults with cerebral palsy. Dev Med Child Neurol 2006, 48:90–95.19. Statistics Canada: Life expectancy, abridged life table, at birth and at age 65,by sex, Canada, provinces and territories, annual (years) (CANSIM Table102–0512). Ottawa: Statistics Canada; 2010.20. Horsman J, Furlong W, Feeny W, Torrence G: The Health Utilities Index(HUI): concepts, measurement properties and applications. Health QualLife Outcomes 2003, 1:54.21. Pogany L, Barr RD, Shaw A, Speechley KN, Barrera M, Maunsell E: Healthstatus in survivors of cancer in childhood and adolescence. Qual Life Res2006, 15:143–157.22. Fryback DG, Dunham NC, Palta M, Hanmer J, Buechner J, Cherepanov D,Herrington SA, Hays RD, Kaplan RM, Ganiats TG, Feeny D, Kind P: US normsfor six generic health-related quality-of-life indexes from the NationalHealth Measurement study. Med Care 2007, 45:1162–1170.23. The Magpie Trial Collaborative Group: Do women with pre-eclampsia, andtheir babies, benefit from magnesium sulphate? The Magpie Trial: arandomised placebo-controlled trial. Lancet 2002, 359:1877–1890.24. British Columbia Ministry of Health: Medical Services Commission paymentschedule. Vancouver: British Columbia Ministry of Health; 2011.25. Ali A, Walentik C, Mantych GJ, Sadiq HF, Keenan WJ, Noguchi A: Iatrogenicacute hypermagnesemia after total parenteral nutrition infusionmimicking septic shock syndrome: two case reports. Pediatrics 2003,112:e70–72.26. Lipsitz PJ: The clinical and biochemical effects of excess magnesium inthe newborn. Pediatrics 1971, 47:501–509.27. Johnson LH, Mapp DC, Rouse DJ, Spong CY, Mercer BM, Leveno KJ, VarnerMW, Iams JD, Sorokin Y, Ramin SM, Miodovnik M, O’Sullivan MJ, PeacemanAM, Caritis SN, Eunice Kennedy Shriver National Institute of Child Healthand Human Development Maternal-Fetal Medicine Units Network:Association of cord blood magnesium concentration and neonatalresuscitation. J Pediatr 2012, 160:573–577.28. Lee S, Anderson L: BC Perinatal services costing project: report on costs in theneonatal intensive care unit. Vancouver: British Columbia Research Institutefor Children’s and Women’s Health; 2004.29. Briggs A, Sculpher M, Claxton C: Decision modelling for health economicevaluation. Oxford: Oxford University Press; 2006.30. Drummond MF, Sculpher MJ, Torrance GW, O’Brien BJ, Stoddart GL: Methodsfor the economic evaluation of health care programmes. 3rd edition. Oxford:Oxford University Press; 2005.31. Tan-Torres E, Baltussen R, Adam T, Hutubessy R, Acharya A, Evans DB,Murray CJ: Making choices in health: WHO guide to cost-effectiveness analysis.Geneva: World Health Organization; 2003.32. National Institute for Health and Clinical Excellence: Briefing paper for theMethods Working Party on the cost effectiveness threshold. London: NationalInstitute for Health and Clinical Excellence; 2007.33. Huusom LD, Secher NJ, Pryds O, Whitfield K, Gluud C, Brok J: Antenatalmagnesium sulphate may prevent cerebral palsy in preterm infants-butare we convinced? Evaluation of an apparently conclusive meta-analysiswith trial sequential analysis. BJOG 2011, 118:1–5.34. Wolf HT, Hegaard HK, Greisen G, Huusom L, Hedegaard M: Treatment withmagnesium sulphate in pre-term birth: a systematic review and meta-analysisof observational studies. J Obstet Gynaecol 2012, 32:135–140.35. Cahill AG, Odibo AO, Stout MJ, Grobman WA, Macones GA, Caughey AB:Magnesium sulfate therapy for the prevention of cerebral palsy inpreterm infants: a decision-analytic and economic analysis. Am J ObstetGynecol 2011, 205:542.e1–7.doi:10.1186/1472-6963-13-527Cite this article as: Bickford et al.: Magnesium sulphate for fetalneuroprotection: a cost-effectiveness analysis. BMC Health Services Re-search 2013 13:527.Submit your next manuscript to BioMed Centraland take full advantage of: • Convenient online submission• Thorough peer review• No space constraints or color figure charges• Immediate publication on acceptance• Inclusion in PubMed, CAS, Scopus and Google Scholar• Research which is freely available for redistributionSubmit your manuscript at www.biomedcentral.com/submitBickford et al. BMC Health Services Research 2013, 13:527 Page 12 of 12http://www.biomedcentral.com/1472-6963/13/527

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.52383.1-0221432/manifest

Comment

Related Items