ECTOPIC PREGNANCY - gyncph.dk pathway, cost effectivnes... · Care pathways for ectopic pregnancy:...

Click here to load reader

  • date post

    02-Apr-2018
  • Category

    Documents

  • view

    218
  • download

    1

Embed Size (px)

Transcript of ECTOPIC PREGNANCY - gyncph.dk pathway, cost effectivnes... · Care pathways for ectopic pregnancy:...

  • AiaaTmifr

    R

    S

    R

    ECTOPIC PREGNANCY

    0d

    Care pathways for ectopic pregnancy: a population-based cost-effectiveness analysisValrie Seror, Ph.D.,a Florent Gelfucci, M.Sc.,a Laurent Gerbaud, M.D., Ph.D.,b

    Jean-Luc Pouly, M.D.,c Herv Fernandez, M.D.,d Nadine Job-Spira, M.D.,e Jean Bouyer, Ph.D.,e

    and Jol Coste, M.D., Ph.D.e

    a INSERM Unit 379, Institut Paoli-Calmettes, Marseille; b Service pidmiologie, prvention et conomie de la sant, CHUClermont-Ferrand, Clermont-Ferrand; c Service de gynco-obsttrique, CHU Clermont-Ferrand, Clermont-Ferrand; d Service degynco-obsttrique, Hpital Antoine Bclre, Clamart; and e INSERM Unit 569, Hpital de Bictre, Le Kremlin-Bictre,France

    Objective: To define care pathways in terms of frequency, costs, and outcomes and to assess their cost-effectiveness.Design: Population-based cost-effectiveness study.Setting: Auvergne EP registry (France).Patient(s): Women (n 1,664) registered between 1994 and 2003.Intervention(s): Standard diagnosis and treatment of EP.Main Outcome Measure(s): Costs before, during, and after hospitalization were assessed from data concerningmedical costs of examinations and treatments. One-year fertility was used for effectiveness assessment. Weassessed cost-effectiveness for the healthcare system.Result(s): Diagnostic ultrasound (47% of scans were nondiagnostic) was essential for the use of methotrexate asa first-line treatment for subacute EP. Hospital and ambulatory care costs were similar for all surgical-carepathways (diagnostic or nondiagnostic ultrasound scan followed by conservative or radical laparoscopy). Hospitaland ambulatory-care costs associated with methotrexate treatment were less than half those for surgical-carepathways. In subacute cases, conservative treatments, and methotrexate in particular, were associated with betterfertility at similar or lower cost to salpingectomy for EP for reproductive failure.Conclusion(s): Conservative treatments are cost-effective with respect to salpingectomy, when subsequentfertility is at stake. Efforts should be made to increase the frequency of diagnostic ultrasound scans, making itpossible to increase methotrexate use and cost-effectiveness. (Fertil Steril 2007;87:73748. 2007 by AmericanSociety for Reproductive Medicine.)

    Key Words: Pregnancy, ectopic, cost, care, cost-effectiveness

    pfdtHaoacwwstssp

    fter a period of decline in industrialized countries, thencidence of ectopic pregnancy (EP) has now stabilized atround 1.5% of pregnancies, or has slightly increased, due tougmented rate of EP resulting from reproductive failure (13).his slight increase (accounted for by recent increases in chla-ydial infections or in smoking in women of reproductive age)

    s worrying because this type of EP has serious consequencesor the subsequent fertility of the woman (infertility in 40% andecurrence in 30% of cases at 2 years) (4).

    eceived February 27, 2006; revised October 31, 2006; accepted No-vember 2, 2006.

    upported by the National Register Committee (Comit National desRegistres, INSERM, InVS), Paris, France.

    eprint requests: Jol Coste, M.D., Ph.D., Dpartement de Biostatistique,Pavillon Saint-Jacques, Hpital Cochin, 27, rue du faubourg Saint-Jacques, 75679 Paris Cedex 14, France (FAX: 33-1-58-41-19-61; E-

    dmail: [email protected]).

    015-0282/07/$32.00oi:10.1016/j.fertnstert.2006.11.005 Copyright 2007 American Soc

    Ectopic-pregnancy care also poses complex questions. Im-rovements in diagnostic methods over the last 3 decades haveacilitated earlier diagnosis (before rupture), leading to theevelopment of new surgical and nonsurgical procedures forhe conservative management of EP and the Fallopian tube (5).owever, despite a large number of noncontrolled studies andmuch smaller number of controlled studies comparing the

    utcomes (immediate success, middle-term fertility) of variouslternatives (6), it remains unclear whether treatment should beonservative (salpingostomy) or radical (salpingectomy),hether conservative management should be drug based, andhether surgical intervention should be laparotomic or laparo-

    copic (7). Economic issues are also of increasing relevance inhe context of limited resources and rising concern withinociety that increases in expenditure should be justified byignificant health benefits. However, few studies have com-ared the costs and cost-effectiveness of different care proce-

    ures for EP (814).

    737Fertility and Sterility Vol. 87, No. 4, April 2007iety for Reproductive Medicine, Published by Elsevier Inc.

  • rFoccmmasst1cehmtie

    MSTci1ddarycrc

    frrtnhebpniFcdmt

    ll

    badesatclwt

    DWppaultcbaa

    ATpdiocipa

    dctrdpbd7skbd

    via

    This study, based on the data from the population-basedegistry of EPs in the Auvergne region (in the center ofrance), aimed to define and assess care pathways in termsf frequency, costs, and outcomes and to evaluate theirost-effectiveness in situations in which the patients clinicalharacteristics were compatible with several different treat-ent options for EP. Data from the registry concerningedical costs of examinations and treatments before, during,

    nd after hospitalization were used to evaluate the wholeequence of care costs, from the first suspicion of EP toubsequent follow-up and possible future pregnancy, fromhe point of view of the healthcare system. We considered-year fertility to be the most relevant outcome for assessingare pathways, and we distinguished acute EP (requiringmergency care as a result of tubal rupture and abundantemoperitoneum) from classical subacute EP because theanagement procedures (and pathways) clearly differ be-

    ween these two types of EP (conservative drug treatment isnappropriate for EP with abundant hemorrhaging, forxample).

    ATERIALS AND METHODStudy Population and Data Collected in the Registryhis study was based on data from the Auvergne EP registryollected from July 1994 to December 2003. The character-stics of this EP register have been described elsewhere (15,6). Briefly, all women aged 1544 years, resident in threepartements (administrative units: Cantal, Allier, and Puye Dme) in the Auvergne Region in the center of France,nd who were treated for EP in one of the area centers wereegistered and prospectively followed until the age of 45ears, and their reproductive outcomes were studied. At eachenter, a trained investigator (a midwife or a physician) isesponsible for case identification, follow-up, and dataollection.

    The information collected for each woman includes theollowing: sociodemographic characteristics and place ofesidence; gynecological, reproductive, and surgical histo-ies; conditions of conception; smoking habits; characteris-ics of the EP (site, rupture, hemoperitoneum); and all diag-ostic and therapeutic procedures (and their order) duringospitalization. For follow-up, the women are interviewedvery 6 months by telephone about whether they wish toecome pregnant again, whether they manage to becomeregnant again, time to pregnancy (or time at risk of preg-ancy), use of contraception or medical measures related tonfertility, and obstetric outcome (if pregnancy occurred).rom April 1999 onward, additional data concerning health-are visits (general practitioners, specialists, and nurses),iagnostic procedures (laboratory and imaging), and treat-ents before and after hospitalization also were prospec-

    ively collected.

    During hospitalization, data were collected for all radio-ogical examinations (including vaginal ultrasound scans);

    aboratory tests (including -hCG, liver tests, and complete i

    738 Seror et al. Cost-effectiveness of ectopic pregnancy ca

    lood-cell count); drug consumption (including antibioticsnd anticoagulants); blood transfusions; therapeutic proce-ures, whether drug-based (methotrexate) or surgical (forxample, conservative, i.e., salpingostomy, or radical, i.e.,alpingectomy); and duration of hospital stay. Before andfter hospitalization, data were collected concerning consul-ations with general practitioners and gynecologists, nursingare, drugs, laboratory tests and ultrasound scans, and sickeave. The data concerning medical care after hospitalizationere collected by means of a questionnaire that was mailed

    o the women 2 months after discharge.

    efinition of Care Pathwayse defined the hospital care pathways that are followed by

    atients, on the basis of the diagnostic- and therapeutic-rocedure data from the registry. Pathways were representeds a tree, with each patient entering at the same point:ltrasound examination. Hospital care pathways were ana-yzed according to the need for emergency care as a result ofubal rupture and abundant hemoperitoneum: a woman wasonsidered to have acute EP if she required resuscitationefore treatment or had an intra-abdominal hemorrhage plusmaximum of 1 day between the first -hCG determination

    nd treatment.

    ssessment of Costs Associated With Care Pathwayshe costs associated with EP care were assessed from theoint of view of the healthcare system. For the assessment ofirect coststhose directly attributable to EP careadopt-ng the healthcare system perspective led us to concentraten hospital and ambulatory costs. The assessment of indirectosts, relating primarily to losses of productivity in patientsn active employment, was based on the daily allowancesrovided by the healthcare system (the national health insur-nce system) for working days lost as a result of EP.

    On the basis of the data provided by the registry for theiagnostic and therapeutic procedures performed for EPare, hospital care costs were calculated by using the na-ional hospital-cost database, giving estimates that were rep-esentative at the national level. The national hospital-costatabase for 2005, based on a sample of 50 public andrivate nonprofit hospitals (3.4 million hospital stays; data-ase available at http://www.atih.sante.fr), uses accountingata to estimate the cost per diagnosis-related group (DRG;71 DRG in total). Because the ectopic-pregnancies DRG ispecific to this condition, grouping together all the differentinds of EP, estimations of hospital costs were free of theiases that may occur when the DRG is a composite index ofiseases.

    The costs of diagnostic and therapeutic procedures in-olved in hospital care for EP were calculated by determin-ng DRG unit values of indices (B for laboratory procedures,nd RCI for relative cost indices relating to surgical and

    maging procedures), with the number of B or RCI involved

    re pathways Vol. 87, No. 4, April 2007

    http://www.atih.sante.fr

  • iltwcwpm(pp

    seottisicdeo

    stesetfpttap

    wsadchei

    wtceacst

    fp

    APStwnwttwWwdt

    ATcswtcdnflrtpa

    ccf

    RCBwwicrtrrct

    aw

    F

    n each procedure taken from the Medical Procedure Cata-ogue (for more information, see the French guidelines forhe economic evaluation of healthcare technologies at http://ww.ces-asso.org [English version available]). Information

    oncerning the drugs administered during hospitalizationas available from the registry (data were available for therincipal drugs administered: antibiotics, anticoagulants, andethotrexate) and from the national hospital-cost database

    mean cost of drugs per hospitalization in the ectopic-regnancies DRG represented 0.96% of the total mean coster stay).

    Data on methotrexate, in particular, were required to as-ess the cost-effectiveness of this treatment option. How-ver, partial information on drugs from the registry led to theverestimation of this component of the cost of hospitaliza-ion because of the addition of the costs of drugs reported inhe registry to the mean cost of drugs per day of hospital-zation derived from the national hospital-cost database. Thislight overestimation of the costs of drugs (a 3% differencen cost with respect to underestimation because of the ex-lusion of drug costs derived from the national hospital-costatabase) had no impact on the outcome of the cost-ffectiveness analysis, because it concerned all the careptions considered to a similar extent.

    The national hospital-cost database was also used to as-ess costs not directly attributable to the diagnostic andherapeutic procedures reported in the registry but that nev-rtheless contributed to the cost of hospital care, such asalaries (nurses, etc.), medical consumables, depreciation ofquipment, and logistics costs (laundry, catering, adminis-rative staff, etc.). These service-center costs can be obtainedrom the national hospital-cost database. They were ex-ressed as mean service-center cost per day of hospitaliza-ion in the ectopic-pregnancies DRG, and their contributiono the costs associated with hospital care pathways was thenssessed on the basis of the duration of hospitalization re-orted in the registry.

    Medical costs incurred before and after hospitalizationere also assessed from the point of view of the healthcare

    ystem. This assessment was based on the fees-for-servicend tariffs of the Common Classification of Medical Proce-ures (http://www.atih.sante.fr). Costs related to sick leaveaused by EP (hospital stay and sick leave before and afterospitalization) were evaluated on the basis of the 41.26-uros daily allowance provided by the French national healthnsurance system.

    Total direct medical costs (hospital and ambulatory care)ere calculated taking into account all medical costs relating

    o EP care both within and outside the hospital. The totalosts of hospitalization were calculated as the sum of costs atach step of the care pathway (diagnosis, first-line treatment,nd possibly, second-line treatment). For care pathways in-luding first-line treatment failure, the costs associated withecond-line treatments were included in the calculation of

    he total cost of the care pathway, taking into account the n

    ertility and Sterility

    requency of first-line treatment failure. All costs were ex-ressed in euros, based on 2005 tariffs.

    ssessment of Subsequent Fertility According to Careathway Followedubsequent fertility after EP treatment was considered to be

    he relevant outcome associated with care pathways. Fertilityas measured as the frequency of women becoming preg-ant (intrauterine pregnancy) during a 1-year period inhich they were trying to conceive. Because the length of

    ime between EP and first attempt to get pregnant was likelyo vary between women, fertility was assessed in the womenho began trying to get pregnant during the 2 years after EP.omen aged younger than 18 or older than 45 years, womenho had been sterilized, and those who declared that theyid not wish to have any more children were excluded fromhe analysis.

    ssessment of the Cost-Effectiveness of Care Pathwayshe cost-effectiveness analysis was designed to compare EPare pathways in terms of costs to the healthcare system andubsequent fertility: we assumed that 1,000 hypotheticalomen entered each of the care pathways studied (metho-

    rexate or laparoscopy performed conservatively or radi-ally). We assumed that all the women tried to get pregnanturing the 2 years after treatment for EP. The cost-effective-ess analysis therefore involved comparing total costs (re-ecting the resources allocated to EP care, given that theseesources could have been allocated for other purposes), withhe total expected outcome of this allocation (intrauterineregnancies). A given first-line treatment was preferred tonother if it was more effective at lower cost.

    Statistical analysis was based on exact 2 tests and Wil-oxon tests on the ranks of individual costs, with 95%onfidence intervals when appropriate (i.e., when the dataollowed a normal or binomial distribution).

    ESULTSare Pathwaysetween July 1, 1994 and December 31, 2003, 1,808 womenere included in the EP registry. Diagnosis or treatment dataere missing for 145 women, so only 1,664 women were

    ncluded in the care-pathways analysis. Overall, 500 EPases resulted from contraceptive failure and 1,164 casesesulted from reproductive failure. Acute EP (as defined inhe paragraph immediately preceding the start of the Mate-ials and Methods section) accounted for 323 cases, with theemaining 1,341 EP cases being considered subacute. Acuteases resulted more frequently from contraceptive failurehan from reproductive failure (24% vs. 17%, P.002).

    Hospital care pathways were defined for all cases of acutend subacute EP (Figs. 1 and 2). All care pathways beganith ultrasound examination (whether the results were diag-

    ostic or not) and were characterized by the first-line treat-

    739

    http://www.ces-asso.orghttp://www.ces-asso.orghttp://www.atih.sante.fr

  • FIGURE 1

    Hospital care pathways for acute EP (n 323). The tree depicts the course of diagnosis and treatment for the patients. Thedecision nodes, generally represented as squares, represent medical choices between surgical and medical procedureswhereas the chance nodes, shown as circles, represent a point in time at which more than one possible event may occur.For example, conservative laparoscopy may have resulted in successful care for EP but, when unsuccessful, further drug-based or surgical treatment had to be initiated. The probability of patients progressing down a given pathway at a chancenode was determined from an outcome frequency obtained from the registry data.

    Seror. Cost-effectiveness of ectopic pregnancy care pathways. Fertil Steril 2007.

    740Seror

    etal.Cost-effectiveness

    ofectopicpregnancy

    carepathw

    aysVol.87,No.4,April2007

  • FIGURE 2

    Hospital care pathways for subacute EP (n 1,341). The tree depicts the course of diagnosis and treatment for the patients.The decision nodes, generally represented as squares, represent medical choices between surgical and medical procedures,whereas the chance nodes, shown as circles, represent a point in time at which more than one possible event may occur.

    Seror. Cost-effectiveness of ectopic pregnancy care pathways. Fertil Steril 2007.

    741Fertility

    andSterility

  • TABLE 1Hospitalization costs in Euros, according to care pathways for subacute and acute ectopic pregnancy (N 751).

    Carepathway Ultrasound

    -hCGdetermination

    Otherlaboratory

    investigationsa Curettage1st-line

    treatment2nd-line

    treatmentb

    Drugs(apartfromMTX)

    Othermedicalcostsc

    Logisticcostsc

    Hospital costper patient

    Subacute ectopic pregnancies (n 604)CLDU

    Median 45.5 9.7 97.1 0 608 0 4.1 385.6 216.9 1,615.2Range 0227.5 058.4 33.1357.3 0288 608608 0608 04.1 01,349.5 0759 686.22,983

    CLnDUMedian 45.5 9.7 129.3 0 608 0 0 371.4 208.9 1,528.3Range 0273 048.7 29.2371 0144 592608 0608 04.1 01,857.1 01,044.5 682.83,678.1

    RLDUMedian 45.5 9.7 101 0 464 4.1 627.3 352.8 1,616.7Range 0182 039 22.4294.3 0288 464592 04.1 01,463.7 0823.2 576.73,150.1

    RLnDUMedian 45.5 9.7 119.5 0 464 4.1 627.3 352.8 1,632.7Range 0182 039 33.1295.3 0288 464592 04.1 02,091 01,176 516.94,186.4

    MTXnDU

    Median 91 9.7 104.9 0 4.9 0 0 126.5 71.2 525.2Range 0364 087.7 0508.5 0288 4.94.9 01,056.1 04.1 0875.1 0492.1 27.32,197.0

    Acute ectopic pregnancies (n 147)CLDU

    Median 45.5 9.74 80.85 72 608 0 0 405.1 227.9 1,535.7Range 0136.5 019.5 11.7165 0288 592608 04.9 04.1 32.71,012.8 18.4569.6 992.72,513.7

    CLnDUMedian 45.5 9.7 119.5 0 608 0 4.1 589.3 331.4 1,717.4Range 0136.5 029.2 41.9239.8 0144 608608 04.9 04.1 01,178.6 0662.8 727.82,648.1

    RLDUMedian 45.5 9.7 96.3 0 464 0 4.13 627.3 352.8 1,599.8Range 0182 029.2 33.1296.5 0144 464592 0464 04.1 01,463.7 0823.2 632.23,178.5

    RLnDUMedian 45.5 9.7 126.8 0 464 4.1 627.3 352.8 1,678.8Range 091 019.5 34.1231.5 0144 464592 04.1 01,672.8 0940.8 642.93,381.6

    Note: CL conservative laparoscopy; CR radical laparoscopy; MTX methotrexate; DU diagnostic ultrasound; nDU nondiagnostic ultrasound.a Other laboratory investigations included blood-cell count, ionogram, C-reactive protein, blood-group determination, and so on.b Average costs of second-line treatments according to the frequencies of first-line treatment failures.c According to the national hospital-cost database from the French DRG program.

    Seror. Cost-effectiveness of ectopic pregnancy care pathways. Fertil Steril 2007.

    742Seror

    etal.Cost-effectiveness

    ofectopicpregnancy

    carepathw

    aysVol.87,No.4,April2007

  • TABLE 2Outpatient costs per patient in Euros, according to care pathways for subacute and acute ectopic pregnancy.

    Outpatient care before hospitalization Outpatient care after hospitalization Sick leaveb

    Carepathway Consultations Ultrasound

    -hCGdetermination

    Cost perpatient Consultations Ultrasound

    -hCGdetermination

    Otherlaboratory

    investigationsaNursing

    careCost perpatient

    No. ofdays

    Cost perpatient

    Subacute ectopic pregnancies: (n 604 before hospitalization; n 315 after hospitalization)CLDU

    Median 23 0 13.5 56.5 21.2 0 12.4 0 0 84.5 10 412.6Range 0129 0283.5 0121.5 0463 0267.4 0209.1 099.6 027.4 0135.4 0335.9 079 03,259.5

    CLnDUMedian 23 0 13.5 59.8 20.4 0 12 0 0 56.4 3.5 144.4Range 0241 0226.8 094.5 0402.6 0135 0201.4 095.9 036 0123.5 0556.5 076.5 03,156.4

    RLDUMedian 23 0 13.5 59.5 23 0 13.5 0 0 50 5 206.3Range 0169 0283.5 067.5 0379 0161 0170.1 054 027 0146.8 0344.4 072 02,970.7

    RLnDUMedian 23 0 0 46 23 0 13.5 0 0 83.3 18 742.7Range 0138 0170 054 0282.4 086 0113.4 054 031 0146.9 0242.8 0163 06,725.4

    MTXnDUMedian 46 0 13.5 77 16 17.2 12.3 0 0 65.5 4 165Range 0184 0453.6 0108 0637.6 0112.3 0158.2 075.3 013.2 066 0330.7 027.5 01,134.7

    Acute ectopic pregnancy: (n 147 before hospitalization; n 91 after hospitalization)CLDU

    Median 23 0 0 23 22.3 0 13.1 0 0 92.8 14.5 599.6Range 069 056.7 040.5 0152.7 2.8102.7 0109.9 0104.6 018.3 0134.7 8.2292.8 154 41.262,228

    CLnDUMedian 23 0 0 43 21.6 0 12.7 0 0 79.7 4.5 185.7Range 086 056.7 040.5 0110.2 064.8 0106.5 076.1 025.4 096.2 0298 224.5 82.521,010.9

    RLDUMedian 23 0 0 23 23 0 13.5 0 0 79.7 8 330.1Range 0103 056.7 013.5 0156.7 0106 0113.4 094.5 031.1 0140.5 0285.4 097 04,002.2

    RLnDUMedian 23 0 0 23 43 0 0 0 0 79.7 12 495.1Range 0132 0283.5 040.5 0413 0132 0113.4 054 027 0140.5 0414.9 253 82.52,186.8

    Note: CL conservative laparoscopy; CR radical laparoscopy; MTX methotrexate; DU diagnostic ultrasound; nDU nondiagnostic ultrasound.a Other laboratory investigations included blood cell count, ionogram, C-reactive protein, Chlamydia trachomatis serological tests, and so on.b Sick leave before, during, and after hospitalization.

    Seror. Cost-effectiveness of ectopic pregnancy care pathways. Fertil Steril 2007.

    743Fertility

    andSterility

  • TABLE 3Total costs of care per ectopic pregnancy according to hospital-care pathway.

    Care pathwayaHospitalcosts ()

    Hospital and outpatientcare costs ()

    Total costs, includingsick leave ()

    Subacute ectopic pregnancies (n 315)CLDU

    Median 1,616.2 1,784.1 2,195.9Range 686.22,983 800.63,393.3 884.36,664.9Mean 1,639.1 1,821.9 2,376.9

    CLnDUMedian 1,510.1 1,671 2,061.2Range 698.63,678.1 833.43,966.5 833.45,278.2Mean 1,609.1 1,791.7 2,196.5

    CL (DU and nDU)Median 1,603.7 1,740.2 2,144.7Range 686.23,678.1 800.63,966.5 833.46,664.9Mean 1,623.5 1,806.3 2,283.3

    RLDUMedian 1,626.7 1,806.2 2,104.7Range 576.73,063.7 691.23,226.9 816.64,925.2Mean 1,546.6 1,722.7 2,284.2

    RLnDUMedian 1,685.3 1,854.7 2,516.3Range 616.64,186.4 716.64,327.4 716.69,208.7Mean 1,687.3 1,859 2,691.1

    RL (DU and nDU)MedianRange 1,656.6 1,819.7 2,200.6Mean 576.74,186.4

    1,610.1691.24,327.4 1,784.3 716.69,208.7 2,468

    MTXDUMedian 454.5 670.5 1,006.3Range 27.32,197 69.52,399.9 69.52,826Mean 643.1 834.7 1,083.8

    Acute ectopic pregnancies (n 91)CLDU

    Median 1,648.2 1,765.8 2,568.4Range 992.72,513.7 1,043.92,686.5 1,082.64,832.7Mean 1,755.2 1,904.7 2,600.4

    CLnDUMedian 1,717.4 1,829.3 2,265.7Range 727.82,648.1 750.82,848.6 1,370.93,701.3Mean 1,745 1,892.6 2,310.1

    CL (DU and nDU)Median 1,717.4 1,829.3 2,400.7Range 727.82,648.1 750.82,848.6 1,082.64,832.7Mean 1,749.7 1,898.2 2,444.1

    RLDUMedian 1,572.7 1,680 2,565.6Range 632.23,178.5 718.23,317.5 718.25,077.3Mean 1,659.9 1,787.6 2,532.8

    Seror. Cost-effectiveness of ectopic pregnancy care pathways. Fertil Steril 2007.

    744 Seror et al. Cost-effectiveness of ectopic pregnancy care pathways Vol. 87, No. 4, April 2007

  • miwrnw

    tvwotclnmu

    adftw

    goccumr

    CCs

    acnat(Efmescirw

    E(adl1ctwftvwu

    twht

    F

    ent provided. For subacute EP, five main pathways weredentified, corresponding to 97% of cases: first-line treatmentas most often methotrexate or conservative or radical lapa-

    oscopy in cases in which ultrasound examination was diag-ostic or was conservative or radical laparoscopy in cases inhich ultrasound examination was not diagnostic.

    For acute EP, four pathways accounted for 85% of cases:he first-line treatment was laparoscopy, which was conser-ative or radical with the care pathways depending onhether the ultrasound scan was diagnostic. The frequencyf nondiagnostic ultrasound scans did not differ according tohe acute or subacute status of the EP (46% vs. 47%), and thehoice of surgical first-line treatment (conservative or radicalaparoscopy) did not depend on whether ultrasound exami-ations were diagnostic or not. However, laparoscopy wasore frequently conservative for subacute EP, regardless of

    ltrasound results (P.003).

    The use of methotrexate as a first-line treatment for sub-cute EP depended on the initial ultrasound scan beingiagnostic. Consistent with this, our data showed that therequencies of both diagnostic ultrasound examinations andhe use of methotrexate as a first-line treatment increasedith time (P.0001 for trend).

    No difference in first-line treatment according to the ori-in of the EP (contraceptive or reproductive failure) wasbserved for acute EP, regardless of ultrasound results. Inontrast, first-line treatment for subacute EP depended on theause of the EP: after the diagnosis of subacute EP byltrasound scan, first-line treatment was more frequentlyethotrexate or conservative laparoscopy for EP caused by

    eproductive failure (71% vs. 59%, P.009).

    osts Associated With EP Care Pathwaysosts incurred before and during hospitalization were as-

    TABLE 3Continued.

    Care pathwayaHospitalcosts ()

    Ho

    RLnDUMedian 1,724.5Range 1,2343,381.6Mean 1,923.2

    RL (DU and nDU)Median 1,626.9Range 632.23,381.6Mean 1,756.1

    Note: CL conservative laparoscopy; CR radical lapanDU nondiagnostic ultrasound.

    Seror. Cost-effectiveness of ectopic pregnancy care pathways. Fertil Steril 2007.

    essed by using data from 751 women who were included P

    ertility and Sterility

    fter April 1, 1999. Costs after hospitalization were finallyalculated for only 406 women, as a result of questionnairesot being returned and missing data. The subsamples of 751nd 406 EP cases used to assess costs were representative ofhe whole sample of 1,664 EP cases in terms of EP severityproportions of acute and subacute EP cases); the origin ofP (proportions of reproductive and contraceptive failure);

    requency of nondiagnostic ultrasound scans; first-line treat-ent failure rates; and the frequencies of all care pathways

    xcept for methotrexate for subacute EP care, which waslightly overrepresented in the subsamples used to assessare costs (19% vs. 14%, P.005). Note that the recentncrease in the use of methotrexate had no impact on theesults of the cost-effectiveness analysis, as each care path-ay was assessed independently.

    Costs were calculated separately for acute and subacuteP care and are presented in Table 1 (hospital care), Table 2

    ambulatory care), and Table 3 (total costs). Hospital costsssociated with surgical-care pathways (diagnostic or non-iagnostic ultrasound followed by conservative or radicalaparoscopy) were largely similar (P.70), ranging from,500 to 1,700 euros per patient. In contrast, the hospitalosts associated with the drug-based care pathway (metho-rexate) were less than half those of the surgical-care path-ays (P.0001). However, for subacute EP care, costs be-

    ore hospitalization were highest for women subsequentlyreated with methotrexate (median cost, 77 vs. 58 euros; Palue for the exact Wilcoxon test 0.0008), because theseomen consulted their doctors more often (P.0004) andnderwent more -hCG determinations (P.0001).

    Pre-hospitalization costs did not differ according to theype of surgical treatment subsequently provided to womenith acute EP (or subacute EP), but ambulatory costs beforeospitalization were lower for women with acute EP than forhose with subacute EP (median cost, 23 vs. 57 euros;

    tal and outpatientare costs ()

    Total costs, includingsick leave ()

    1,918.4 2,552.4,4283,673.9 1,623.45,579.8

    2,105.3 2,752.5

    1,782.6 2,55918.23,673.9 718.25,579.8

    1,903.7 2,613.1opy; MTX methotrexate; DU diagnostic ultrasound;

    spic

    1

    7

    rosc

    .0001): women with acute EP consulted their doctors less

    745

  • fs(cEta1u(

    awcwacab

    ouipoh

    FFwfgnn(ari

    pmgiw(lrrTn

    CPOc

    ABLE

    4S

    ubse

    que

    ntfe

    rtili

    tya

    acco

    rdin

    gto

    hosp

    ital

    care

    pat

    hway

    for

    sub

    acut

    eec

    top

    icp

    reg

    nanc

    ies

    (n

    517)

    .

    Car

    ep

    athw

    ay:

    ultr

    aso

    und

    cate

    go

    ryb

    yfi

    rst-

    line

    trea

    tmen

    tN

    oat

    tem

    pt

    or

    IVF

    Att

    emp

    tsto

    get

    pre

    gna

    nt

    No

    pre

    gna

    ncy

    Mis

    carr

    iag

    eo

    rE

    PIn

    trau

    teri

    nep

    reg

    nanc

    ies

    Con

    serv

    ativ

    ela

    par

    osco

    py

    123

    (54.

    2;47

    .76

    0.7)

    Dia

    gnos

    tic(n

    12

    0)22

    (18.

    3;11

    .42

    5.3)

    25(2

    5.5;

    16.9

    34.

    1)21

    (21.

    4;13

    .32

    9.6)

    52(5

    3.1;

    43.2

    62.

    9)N

    ond

    iagn

    ostic

    (n

    169)

    40(2

    3.7;

    17.3

    30.

    1)29

    (22.

    5;15

    .32

    9.7)

    29(2

    2.5;

    15.3

    29.

    7)71

    (55.

    0;46

    .56

    3.6)

    Rad

    ical

    lap

    aros

    cop

    y42

    (44.

    2;34

    .25

    4.2)

    Dia

    gnos

    tic(n

    72

    )30

    (41.

    7;30

    .35

    3.1)

    16(3

    8.1;

    23.4

    52.

    8)9

    (21.

    4;9.

    0)3

    3.8

    17(4

    0.5)

    (25.

    655

    .3)

    Non

    dia

    gnos

    tic(n

    92

    )39

    (42.

    4;32

    .35

    2.5)

    15(2

    8.3;

    16.2

    40.

    4)13

    (24.

    5;12

    .93

    6.1)

    25(4

    7.2;

    33.7

    61.

    6)M

    etho

    trex

    ate:

    dia

    gnos

    tic(n

    64

    )18

    (28.

    1;17

    .13

    9.1)

    10(2

    1.7;

    9.8

    33.7

    )10

    (21.

    7;9.

    833

    .7)

    26(5

    6.5;

    42.2

    70.

    9)N

    ote:

    All

    dat

    aar

    enu

    mb

    er(%

    ;95

    %co

    nfid

    ence

    inte

    rval

    ).a

    Sub

    seq

    uent

    fert

    ility

    asse

    ssed

    atth

    een

    dof

    a1-

    year

    per

    iod

    beg

    inni

    ngd

    urin

    gth

    e2

    year

    saf

    ter

    EP

    care

    .

    Sero

    r.C

    ost-

    effe

    ctiv

    enes

    sof

    ecto

    pic

    preg

    nanc

    yca

    repa

    thw

    ays.

    Fer

    til

    Ster

    il20

    07.

    requently (P.0006) and underwent less frequent ultra-ound examinations (P.0001) and -hCG measurementsP.0001). No difference was found in post-hospitalizationosts according to the type of surgical treatment or type ofP (acute or subacute). However, medical costs after hospi-

    alization were higher if the hospital care pathway for sub-cute EP was drug-based rather than surgical (median cost,30 vs. 80 euros; P.0009), because patients had moreltrasound examinations and -hCG determinationsP.0001).

    Overall, the ambulatory and hospital costs of care forcute or subacute EP ranged from 1,700 euros to 1,900 euroshen the first-line treatment was surgical, whereas these

    osts amounted to about 700 euros for subacute EP treatedith methotrexate. The amount of sick leave taken before

    nd after hospitalization did not differ according to hospitalare pathway or type of EP. In particular, no difference in themount of sick leave taken was observed between drug-ased and surgical-care pathways for subacute EP.

    The cost of treating acute and subacute EP did not dependn the origin of the EP (contraceptive or reproductive fail-re), with no difference in medical costs found before, dur-ng, or after hospitalization, regardless of the hospital careathway considered. The origin of the EP also had no effectn the length of hospital stay or sick leave before and afterospitalization.

    ertility After EP Care and Efficacy of Care Pathwaysertility assessment was based on the data for 630 womenith EP caused by reproductive failure and at least 3 years of

    ollow-up (to ensure a period of at least 1 year of trying toet pregnant for women who began trying for a new preg-ancy in the 2 years after the index EP). This subsample didot differ from the whole sample of reproductive-failure EPn 1,164) in terms of the type of EP (acute or subacute)nd the care pathways. We were unable to assess fertilityates after acute EP care because of the small number ofntrauterine pregnancies per hospital care pathway.

    For subacute EP cases, 20% of women did not try to getregnant again after EP care, and 8% opted for IVF treat-ent. Of the remaining women, 26% did not succeed in

    etting pregnant, and 52% had intrauterine pregnancies lead-ng to the birth of a child (Table 4). Similar fertility ratesere obtained with conservative treatments: methotrexate

    57%; 95% confidence interval, 42%71%) and conservativeaparoscopy (54%; 95% confidence interval, 48%61%);adical laparoscopy (salpingectomy) gave the lowest fertilityate (44%; 95% confidence interval, 34%54%, P.0001).hese rates were not affected by the diagnostic or nondiag-ostic nature of the ultrasound examination before treatment.

    ost-Effectiveness of Subacute Ectopic Pregnancy Careathwaysn the basis of the results presented above, we assessed the

    ost-effectiveness of treatment options for subacute EP re-

    746 Seror et al. Cost-effectiveness of ectopic pregnancy care p

    T

    athways Vol. 87, No. 4, April 2007

  • suphiso

    boiwsdalitme

    DTfmocvaaest

    nkcmttfncdeucrsofrct

    aaclposmmwcaor(tcld

    shrTcfnrmhdsirHa

    amrfsghteypcabp

    F

    ulting from reproduction failure, regardless of whether theltrasound examination was diagnostic or not. For each careathway, the total costs associated with EP care for 1,000ypothetical women (Table 3) and the expected number ofntrauterine pregnancies after EP care (Table 4) were as-essed, making it possible to compare care pathways in termsf costs and expected outcomes.

    First-line treatment with methotrexate was cost-effectiveecause much lower levels of resource allocation (one half tone third) were required to obtain an outcome in terms ofntrauterine pregnancies at least as good as that obtainedith conservative or radical laparoscopy. Total costs were

    imilar for conservative and radical laparoscopy (about 3%ifference), but subsequent fertility appeared to be betterfter conservative laparoscopy, suggesting that conservativeaparoscopy may be cost-effective when compared with rad-cal laparoscopy (salpingectomy). Similar results were ob-ained if mean costs per woman were considered rather thanedian costs, and a sensitivity analysis on costs had no

    ffect on the findings.

    ISCUSSIONhis cost-effectiveness analysis dealt with the care pathways

    ollowed by women presenting EP in the real world ofedical practice, rather than considering pairs of diagnostic

    r therapeutic alternatives for highly selected cases. It fo-used on the womans subsequent fertility as the most rele-ant outcome, rather than on detection or complication rates,s generally used in previous studies (814). This study waslso population based rather than university hospital or ref-rence center based, and it distinguished acute EP fromubacute EP, because care pathways clearly differ for thesewo types of EP.

    By using this approach, we showed that the diagnostic orondiagnostic nature of the initial ultrasound scan was theey component for medical decision-making concerning theare of subacute EP, because in practice, methotrexate treat-ent was only possible if the ultrasound scan was diagnos-

    ic. Our finding that about one in two ultrasound examina-ions is nondiagnostic therefore has important implicationsor the costs of subacute EP care. This high frequency ofondiagnostic ultrasound scans in routine practice is rarelyonsidered in classical diagnostic studies (17) and was sel-om discussed in previous studies comparing the cost-ffectiveness of methotrexate and its alternatives. Otherwise,ltrasound results (diagnostic or not) had no effect on theosts of care and subsequent fertility associated with lapa-oscopy (whether radical or conservative). Care pathways forubacute EPs showed a gradual increase in the use of meth-trexate with time, but the frequency of first-line treatmentailure remained constant, suggesting that decision-makingegarding first-line treatment depends more on medicalhoices between substitutable options than on clinical or

    echnical constraints. t

    ertility and Sterility

    This study clearly shows that all conservative treatments,nd methotrexate in particular, are cost-effective for sub-cute EP resulting from reproductive failure. Indeed, theseonservative treatments gave better outcomes at equal orower cost to the healthcare system. This finding, which wasartly expected from cost studies indicating the superiorityf methotrexate (6, 8, 10, 12, 18, 19) and from fertilitytudies indicating the superiority of all conservative treat-ents (20, 21), has clear consequences in terms of recom-endations for practice. When future fertility is at stake (i.e.,hen EP results from a reproductive failure), in practice,

    onservative treatments should be preferred, with methotrex-te used whenever possible, for subacute cases (about 80%f all cases) that have a positive diagnostic ultrasound (cur-ently only about 50% of subacute cases). Radical treatmentsalpingectomy) was found to be less costly than conserva-ive laparoscopy in a Dutch study (22), but this was not thease in our study. Salpingectomy should henceforth haveimited indications in the context of EP as a result of repro-uctive failure.

    Several methodological aspects of this study merit discus-ion. We tried to maximize the relevance of our data toealth systems outside France by collecting detailed data onesource use within and outside hospitals for each patient.his painstaking, prospective data collection procedure fa-ilitates comparisons with similar studies performed at dif-erent times or in different health systems. The use of theational hospital-cost database provided estimates of costsepresentative of the situation throughout France, unlikeethodologies based on charges or bills that are provided by

    ospital accounting departments (hospitals often have veryifferent accounting practices and budget equilibrium con-traints). However, our study has limitations. First, EP caren the French region of Auvergne cannot be considered to beepresentative of EP care worldwide, or even in Europe.owever, its variety (shown in a previous study [23]) may be

    n advantage in this comparative cost-effectiveness study.

    Second, there were missing data concerning medical costsnd sick leave after hospitalization. It remains possible thatedical costs differed in nature and quantity in a manner

    elated to whether a woman returned her questionnaire onollow-up care. However, analysis of the available datahowed that the impact of such differences would be negli-ible, given the much lower costs of ambulatory than ofospital care. Third, despite the large size of our sample,here were too few acute EP cases for a reliable cost-ffectiveness analysis, and we therefore restricted our anal-sis to the women with subacute EP. Finally, our study wasurely observational. However, randomized trials for all EPare pathways are difficult to perform, or even to advocate,t the regional level. Population-based studies, such as thoseased on morbidity registers, have been shown to be com-lementary to randomized controlled trials for assessing

    herapeutic or health interventions (24).

    747

  • pwsur

    R

    1

    1

    1

    1

    1

    1

    1

    1

    1

    1

    2

    2

    2

    2

    2

    In summary, conservative treatments, and methotrexate inarticular, appeared to be cost-effective for EP cases inhich the preservation of fertility was important. Efforts

    hould be made to decrease the proportion of nondiagnosticltrasounds, which prevent the use of methotrexate andeduce the cost-effectiveness of care.

    EFERENCES1. Centers for Disease Control. Ectopic pregnancyUnited States, 1990

    92. JAMA 1995;273:533.2. Coste J, Bouyer J, Germain E, Ughetto S, Pouly JL. Recent declining

    trend in ectopic pregnancy in France: evidence of two clinicoepidemio-logic entities. Fertil Steril 2000;74:8816.

    3. Coste J, Bouyer J, Ughetto S, Gerbaud L, Fernandez H, Pouly JL, et al.Ectopic pregnancy is again on the increase. Recent trends in theincidence of ectopic pregnancies in France (19922002). Hum Reprod2004;19:20148.

    4. Bernoux A, Job-Spira N, Germain E, Coste J, Bouyer J. Fertilityoutcome after ectopic pregnancy and use of an intrauterine device at thetime of the index ectopic pregnancy. Hum Reprod 2000;15:11737.

    5. Ory SJ. New options for diagnosis and treatment of ectopic pregnancy.JAMA 1992;267:5347.

    6. Hajenius PJ, Mol BW, Bossuyt PM, Ankum WM, Van Der Veen F.Interventions for tubal ectopic pregnancy. Cochrane Database Syst Rev2000;CD000324.

    7. Maymon R, Shulman A. Controversies and problems in the currentmanagement of tubal pregnancy. Hum Reprod Update 1996;2:54151.

    8. Creinin MD, Washington EA. Cost of ectopic pregnancy management:surgery versus methotrexate. Fertil Steril 1993;60:9639.

    9. Gray DT, Thorburn J, Lundorff P, Strandell A, Lindblom B. A cost-effectiveness study of a randomized trial of laparoscopy versus lapa-rotomy for ectopic pregnancy. Lancet 1995;345:113943.

    0. Morlock RJ, Lafata JE, Eisenstein D. Cost-effectiveness of single-dosemethotrexate compared with laparoscopic treatment of ectopic preg-nancy. Obstet Gynecol 2000;95:40712.

    1. Durston WE, Carl ML, Guerra W, Eaton A, Ackerson LM. Ultrasoundavailability in the evaluation of ectopic pregnancy in the ED: compar-ison of quality and cost-effectiveness with different approaches. Am J

    Emerg Med 2000;18:40817.

    748 Seror et al. Cost-effectiveness of ectopic pregnancy ca

    2. Sowter MC, Farquhar CM, Gudex G. An economic evaluation ofsingle dose systemic methotrexate and laparoscopic surgery for thetreatment of unruptured ectopic pregnancy. BJOG 2001;108:204 12.

    3. Gracia CR, Barnhart KT. Diagnosing ectopic pregnancy: decision anal-ysis comparing six strategies. Obstet Gynecol 2001;97:46470.

    4. Ailawadi M, Lorch SA, Barnhart KT. Cost-effectiveness of presump-tively medically treating women at risk for ectopic pregnancy comparedwith first performing a dilatation and curettage. Fertil Steril 2005;83:37682.

    5. Coste J, Job-Spira N, Aublet-Cuvelier B, Germain E, Glowaczover E,Fernandez H, et al. Incidence of ectopic pregnancy. First results of apopulation-based register in France. Hum Reprod 1994;9:7425.

    6. Job-Spira N, Bouyer J, Pouly JL, Germain E, Coste J, Aublet-CuvelierB, et al. Fertility after ectopic pregnancy: first results of a population-based cohort study in France. Hum Reprod 1996;11:99104.

    7. Murray H, Baakdah H, Bardell T, Tulandi T. Diagnosis and treatmentof ectopic pregnancy. CMAJ 2005;173:90512.

    8. Lecuru F, Robin F, Chasset S, Leonard F, Guitti S, Taurelle R. Directcost of single dose methotrexate for unruptured ectopic pregnancy.Prospective comparison with laparoscopy. Eur J Obstet Gynecol Re-prod Biol 2000;88:16.

    9. Vaissade L, Gerbaud L, Pouly JL, Job-Spira N, Bouyer J, Coste J, et al.[Cost-effectiveness analysis of laparoscopic surgery versus methotrex-ate: comparison of data recorded in an ectopic pregnancy registry]. JGynecol Obstet Biol Reprod 2003;32:44758 (in French).

    0. Bouyer J, Job-Spira N, Pouly JL, Coste J, Germain E, Fernandez H.Fertility following radical, conservative-surgical or medical treatmentfor tubal pregnancy: a population-based study. BJOG 2000;107:71421.

    1. Bangsgaard N, Lund CO, Ottesen B, Nilas L. Improved fertility fol-lowing conservative surgical treatment of ectopic pregnancy. BJOG2003;110:76570.

    2. Mol BW, Hajenius PJ, Engelsbel S, Ankum WM, Hemrika DJ, van derVeen F, et al. Is conservative surgery for tubal pregnancy preferable tosalpingectomy? An economic analysis. BJOG 1997;104:8349.

    3. Coste J, Bouyer J, Fernandez H, Pouly JL, Job-Spira N. A population-based analytical approach to assessing patterns, determinants and out-comes of health care, with an application to ectopic pregnancy. MedCare 2000;38:73949.

    4. Hornberger J, Wrone E. When to base clinical policies on observational

    versus randomized trial data. Ann Intern Med 1997;127:697703.

    re pathways Vol. 87, No. 4, April 2007

    Care pathways for ectopic pregnancy: a populationbased cost-effectiveness analysisMATERIALS AND METHODSStudy Population and Data Collected in the RegistryDefinition of Care PathwaysAssessment of Costs Associated With Care PathwaysAssessment of Subsequent Fertility According to Care Pathway FollowedAssessment of the Cost-Effectiveness of Care Pathways

    RESULTSCare PathwaysCosts Associated With EP Care PathwaysFertility After EP Care and Efficacy of Care PathwaysCost-Effectiveness of Subacute Ectopic Pregnancy Care Pathways

    DISCUSSIONREFERENCES