|
Cost-effectiveness of strategies used in the evaluation of pregnancies complicated by elevated maternal serum alpha-fetoprotein levels |
Nadel A S, Norton M E, Wilkins-Haug L |
|
|
Record Status This is a critical abstract of an economic evaluation that meets the criteria for inclusion on NHS EED. Each abstract contains a brief summary of the methods, the results and conclusions followed by a detailed critical assessment on the reliability of the study and the conclusions drawn. Health technology Strategies used in the evaluation of pregnancies with elevated maternal serum alpha-protein levels viz, targeted ultrasound and amniocentesis.
Economic study type Cost-effectiveness analysis.
Study population A hypothetical cohort of 100,000 women with singleton pregnancies who underwent serum screening at 16-18 menstrual weeks and who agreed to obtain further recommended diagnostic evaluation.
Dates to which data relate Data relating to the prevalence of structurally abnormal foetuses, and the sensitivity and specificity of the diagnostic procedures were mostly obtained from literature based estimates, ranging from 1977 to 1995. To estimate the false-positive rate of elevated MSAFP, data from the Antenatal Diagnostic Center between 1st October 1991 and 1st February 1995 were analysed. Hospital charges were used to derive costs but no dates were given for this data.
Source of effectiveness data Data on final outcomes were derived from modelling based on information given in the literature and various assumptions.
Link between effectiveness and cost data The data for the model on the prevalence of abnormalities and the performance of the diagnostic tests were either derived via data retrieval, published sources or from assumptions.
Study sample 7,011 pregnancies recorded on the data base of the Antenatal Diagnostic Center between 1st October 1991 and 1st February 1995. Power calculations were not reported.
Study design Retrospective data retrieval, case series.
Analysis of effectiveness The objective was to find out in how many cases the samples were at 2.0 multiples of the median (MoM) or 2.5 MoM.
Effectiveness results 206 samples (2.9%) were at least 2.0 MoM and 127 samples (1.8%) were at lest 2.5 MoM.
Clinical conclusions This result is in agreement with a recent study which also indicates that false-positive rates associated with MSAFP are improving.
Modelling Modelling was used to compare the six different diagnostic protocols which are possible when women are found to have elevated levels of MSAFP. The model reports on abnormalities identified and the total and incremental cost of these abnormalities. Data on the performance of the diagnostic tests in terms of specificity and sensitivity were either derived from published sources or were the result of assumptions made. This information was applied to a hypothetical cohort of 100,000 women to reach the outcome results for abnormalities identified and their total and incremental costs.
Outcomes assessed in the review No formal review was undertaken but references were given for the data used in the model. At least one reference was given for the following data which were used in the model: prevalence of structural abnormalities associated with elevated MSAFP; sensitivity and specificity of MSAFP;sensitivity of ultrasound for detecting a structural abnormality associated with elevated MSAFP; prevalence of aneuploidy.
Study designs and other criteria for inclusion in the review Sources searched to identify primary studies Criteria used to ensure the validity of primary studies Methods used to judge relevance and validity, and for extracting data Number of primary studies included Methods of combining primary studies Investigation of differences between primary studies Results of the review The data derived from published sources and used in the model were as follows:prevalence of structural anomalies associated with raised levels of MSAFP (in a hypothetical cohort of 100,000 women): neural-tube defect, 70; ventral-wall defect, 40. The sensitivity of MSAFP is 90% for open spina bifida and 83% for ventral wall defects at 2.0 MoM; the sensitivities at 2.5 MoM are 80% and 77% respectively. The sensitivity of ultrasound for detecting a structural abnormality associated with elevated MSAFP was 94%. The prevalence of aneuploidy in cases with an elevated MSAFP was stated to be 1.0%.
Methods used to derive estimates of effectiveness Opinion based estimates of effectiveness for some parameters were formed from the authors' assumptions.
Estimates of effectiveness and key assumptions It was assumed that:
(1) the sensitivity of elevated MSAFP for encephalocele is similar to that for open spinal bifida;
(2) sensitivity of amniotic fluid AFP and acetylcholinesterase for structural abnormalities is 100;
(3) the specificity of both ultrasound and amniocentesis for structural abnormalities is 100%; and
(4) ultrasound is unable to detect foetal aneuploidy.
Measure of benefits used in the economic analysis The outcome measure used was the number of anomalous foetuses identified.
Direct costs The cost perspective was that of a third-party payer whose costs are those of the screening programme. The cost of raising a child with an abnormality was not evaluated, and the non-monetary cost to the mother of iatrogenic foetal loss was not included. The data on costs were derived from Brigham and Women's hospital charges multiplied by 0.4,which corresponds to the average third-party reimbursement rate. The cost of collecting and analysing maternal serum samples or for screening ultrasound were not included because, under consideration here, were the costs of diagnostic examinations once these abnormal values are present. The costs of further diagnostic evaluations were also not included since, as the authors stated, such costs tend to offset each other and represent only a fraction of the total cost. Some costs and quantities were reported separately. The dates for the costs were not reported.
Statistical analysis of costs Sensitivity analysis A one way sensitivity analysis on the sensitivity of ultrasound was performed.
Estimated benefits used in the economic analysis The numbers of abnormalities detected via the 6 protocols in 100,000 women were as follows: Protocol A = 82;Protocol B = 90; Protocol C = 87; Protocol D = 96; Protocol E = 105; Protocol F = 125. In terms of iatrogenic foetal loss, there were no losses in protocols A and B, 9 in Protocol C, 14 in protocol D, 9 in protocol E and 14 losses in protocol F.
Cost results The costs to be borne by third-party payers using the method described earlier are as follows: targeted ultrasound, $176; amniocentesis with measurement of amniotic fluid AFP and acetylcholinesterase, $356; amniocentesis with measurement of amniotic fluid AFP, acetylcholinesterase and foetal karyotyping, $668. These costs each include genetic counselling ($40).
Synthesis of costs and benefits Total cost per abnormality identified was: Protocol A = $3,900; Protocol B = $5,700; Protocol C = $7,400; Protocol D = $10,800; Protocol E = $11,500; and Protocol F =$15,500.
Incremental cost per abnormality identified versus protocol A was: Protocol B = $24,200; Protocol C = $64,800; Protocol D = $51,100; Protocol E = $38,500; and Protocol F=$37,700.
Incremental cost per abnormality identified versus protocol B was: Protocol D = $87,000; Protocol E = $46,100; Protocol F = $40,800.
Varying the sensitivity of ultrasound from 80 to 100% had little effect on the cost-effectiveness of protocol B. The incremental cost of protocol E over protocol B fell as the sensitivity of ultrasound declined but remained above the cost per detected abnormal foetus of the triple panel, even at a sensitivity of only 80%.
Authors' conclusions The authors stated that theywould find it difficult to justify routine use of amniocentesis rather than targeted ultrasound for women with moderate elevation in MSAFP. This is due to the foetal losses and expense associated with amniocentesis and the fact that few additional abnormal foetuses are detected. Further, they stated that since the sensitivity of ultrasound for structural abnormalities associated with elevated MSAFP is so high, using amniocentesis rather than ultrasound as a confirmatory test provides only a small increment in diagnostic yield. The main reason for choosing amniocentesis over targeted ultrasound is to identify foetal aneuploidy. Protocols involving only amniocentesis for detection of AFP and acetylcholinesterase but not karyotyping have been shown not to be cost-effective in this analysis. However, almost half of aneuploidies are sex chromosomal abnormalities which carry a better prognosis and do not so often lead to the choosing of a termination.
CRD COMMENTARY - Selection of comparators The comparators were based on reported current practice.
Validity of estimate of measure of benefit The outcomes of the model were clearly stated although the study would have benefited from a more detailed report of how the input values were obtained, and whether these were representative. Sometimes only one study was quoted as the source of input data and it is impossible to ascertain whether this one study was well carried out and representative. Information on how the assumptions were made would also have increased confidence that the final outcomes are reliable. Failing this a sensitivity analysis on all variables would have been possible, although the authors did perform a small sensitivity analysis on what they considered to be the most important variable.
Validity of estimate of costs Hospital charges were used, multiplied by a reimbursement charge of 0.4 and policy makers will have to consider if this is applicable to their own cost situation. Since the costs used are clearly stated, a comparison can easily be made.
Bibliographic details Nadel A S, Norton M E, Wilkins-Haug L. Cost-effectiveness of strategies used in the evaluation of pregnancies complicated by elevated maternal serum alpha-fetoprotein levels. Obstetrics and Gynecology 1997; 89(5 Part 1): 660-665 Other publications of related interest Comment in: Obstetrics andGynecology 1997;90(2):317-8.
Indexing Status Subject indexing assigned by NLM MeSH Amniocentesis /adverse effects /economics /standards; Congenital Abnormalities /prevention & Cost-Benefit Analysis; Female; Fetal Death /etiology; Humans; Pregnancy; Pregnancy Complications /blood; Prevalence; Reproducibility of Results; Sensitivity and Specificity; Ultrasonography, Prenatal /economics /standards; alpha-Fetoproteins /metabolism; control AccessionNumber 21997000670 Date bibliographic record published 31/03/1999 Date abstract record published 31/03/1999 |
|
|
|