|
Screening primiparous women and newborns for fetal/neonatal alloimmune thrombocytopenia: a prospective comparison of effectiveness and costs |
Durand-Zaleski I, Schlegel N, Blum Boisgard C, Uzan S, Dreyfus M, Kaplan C, Aujard Y, Baumann C, Blot P, Boissinot C, Daffos F, Fernandez H, Forestier F, Hurtaud-Roux F, Oury J F, Pons J C, Tchernia G, Verdy E |
|
|
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 The detection and subsequent treatment of fetomaternal alloimmune thrombocytopenia (FMAIT)/Neonatal alloimmune thrombocytopenia (NAIT). This is a disease caused by the maternal immunisation against foetal platelet antigens which can occur when the mother has the antigen HPA-1b and the father is HPA-1a. The techniques employed were: radioimmunoassay and monoclonal antibody-specific immobilization of platelet antigen.
Economic study type Cost-effectiveness analysis.
Study population Primiparous women were included in the pregnant women screening strategy and newborns in the newborn strategy. Newborns whose mothers were part of the primiparous protocol were also included in the newborn population. No details of the patient populations were given.
Setting The primiparous and newborn strategies were carried out simultaneously in 3 maternity wards. The study and the economic analysis were conducted in France.
Dates to which data relate No dates were provided for the effectiveness study, except that the study was undertaken during a 35 month period. Neither were dates given for the period in which data on resource use were collected. Prices were quoted in either 1993 or 1995 prices.
Source of effectiveness data The following data were derived from a single study: the number of women developing antibodies and the number of foetuses with FMAIT; number of newborns with anti-HPA-1a alloimmunisation.
Link between effectiveness and cost data The costing was undertaken prospectively on the same patient sample.
Study sample No randomisation was used although power calculations were employed to determine sample sizes. According to the power calculations, 1,842 primiparous women would need to be included in order to have a 99% chance of at least one case and 8,985 newborns would be necessary to find at least one case. In the actual study, 2,066 primiparous and 6,081 newborns were included. There were no control groups. 2,066 out of an estimated 3,000 primiparous women presenting with pregnancy during this period were included in the study (69%). Of the remainder, 20% refused to participate and 11% were private patients. Out of the 52 HPA-1b women detected, 3 refused to participate further and 2 moved away. During the same period, 6,081 newborns were screened out of 8,836 live births in all three wards (69%). Of those not included in the study, 25% were weekend deliveries and 7% comprised newborns who were born with conditions requiring blood work-up in any event. Of the 6,081 cord samples, 449 (7%) were excluded because of blood clotting or platelet aggregates. Platelet numeration was therefore only performed on 5,632 blood samples (64%).
Study design This was a prospective cohort study in relation to the screening of pregnant women and a large-scale case series in relation to the neonatal screening programme. No randomisation took place. The study was reported to have taken place in 3 maternity wards. The period of follow-up for effectiveness analysis was not stated.
Analysis of effectiveness The analysis was based on treatment completers only. Primary health outcomes: foetal death/disability arising from anti-HPA-1a; women detected with HPA-1a immunisation; detection of HPA-1a alloimmune thrombocytopenia among infants. A random sample (100) of newborns who were screened was compared with a sample of 100 unscreened newborns to test whether there was a difference in the populations screened and not screened. Significant differences at the 5% level were birthweight and maternal age.
Effectiveness results 52/2,066 women were HPA-1b. Of these women, 2 were found to have partners who were HPA-1b which meant the foetus was not at risk of FMAIT. Of the 50 women at risk, four were found to have produced anti-HPA-1a alloantibodies during the study. Of these, two women agreed to foetal blood puncture and two declined. Both the foetuses tested were found to have FMAIT; the other two women delivered normal infants without FMAIT. Of the women found to be carrying a foetus exhibiting FMAIT, one was treated with IVIg and the other received corticosteroids. The first treatment was effective and vaginal delivery was possible and the platelet count favourable. The other woman had an emergency caesarean section due to cord haemorrhage and acute foetal distress during the foetal blood sampling. This infant was anaemic and thrombocytopenic and treated with red blood cells, HPA-1b platelet transfusions and IVIg. In relation to the newborns, 48/6081 were found to be thrombocytopenic with low platelet counts. 5/48 occurred in HPA-1a children whose mothers were HPA-1b and in all 5 mothers anti-HPA-1a antibodies were found (1 was born to a primiparous woman). 4 autoimmunisations and 12 possible allo- and/or auto-immunisations were found. A total of 22 thrombocytopenias resulted from immunisations.
Clinical conclusions In-utero blood sampling may be associated with foetal loss and, in this study, one of the two women agreeing to undergo this procedure suffered serious complications (emergency caesarian section with no foetal loss).
Measure of benefits used in the economic analysis Benefits were as described in the effectiveness analysis using the following assumptions as applied to the general French population: 90% feasibility rate; birth rate of 13.5/1,000; mean number of children per woman 1.8; rate of primiparous women, 46%.
Direct costs The cost perspective was that of the French health care financing administration. Only direct costs were considered. The costs of tests and procedures were reported to have been estimated from the charge to the payer. The cost of treatments was the purchase price, by the hospital, of blood and drug products. Excluded from the cost-effectiveness calculations, but nevertheless presented as indicators of future costs, were the costs of disability. Quantities and costs were analysed separately, the latter including the costs of diagnosis, treatment and complications. In terms of complications, the risk of accident from foetal blood sampling was also computed (assumed to be 3%) and it was further assumed that all foetuses suffering from cord wound were lost. Cost data relating to the cost of screening primiparous women was presented in 1993 dollars, whereas cost data for the screening of newborns was presented in 1995 prices. Part of the benefit associated with screening pregnant women is that the information is useful to the current pregnancy as opposed to newborn screening, which is reported only to provide information for future pregnancies. The cost-effectiveness of the dual strategy was weighted to take account of this.
Statistical analysis of costs No statistical analysis of costs was performed.
Sensitivity analysis A sensitivity analysis was performed on the inclusion of the cost of caesarean sections. This was not included in the main analysis since the occurrences were reported to be too rare. Sensitivity analysis was also performed on the likely number of missed HPA-1b patients who were treated privately and therefore not included in the study. The prevalence of NAIT in the population of unscreened newborns was varied to see if this had any impact on the overall result, as was the threshold chosen to separate "normal" newborns from thrombocytopenic newborns. All sensitivity analyses took the form of one-way simple analyses.
Estimated benefits used in the economic analysis In the screening of primiparous women (2,066), the following number of cases were detected: HPA-1b (52), HPA-1b and father HPA-1a (50), alloimmunisation (4), foetal thrombocytopenia (2). In the screening of newborns (6,081), the following cases were found: Thrombocytopenia(48), anti-HPA-1a (5), allo/auto (12), auto (4). Side-effects were included in the economic analysis: the risk of accident arising from foetal blood sampling was assumed to be 3% and, from these, it was assumed there would be no survivors.
Cost results The total cost of the screening strategy for primiparous women was $198,860, which included diagnosis costs of $182,060 and treatment costs of $16,800 for IVIg corticosteroids (1993 prices). For the screening of newborns the total cost was $93,248 which included a total diagnosis cost of $92,048 and $1,200 for treatment. The cost per million inhabitants screened was $740,000 for the primiparous and $207,000 for the newborn strategy. These latter costs are highly dependent on the age structure and birthrate of the population.
Synthesis of costs and benefits The cost-effectiveness of the primiparous strategy varied from $45,000 per immunization found to $500,000 per death/disability avoided. The cost-effectiveness of the newborn strategy varied from $18,000 per immunization found to $225,000 per death/disability avoided. In the sensitivity analysis, the impact of pregnant women being cared for privately was assessed. It was found that if the number of missed HPA-1b patients were 4 or 1%, the cost-effectiveness ratios of the primiparous and newborn strategies for the detection of HPA-1a alloimmune thrombocytopenia would be equal. Including the savings due to averted caesarean sections, the cost of caesarean sections was also tested in the sensitivity analysis. It was found that if it is assumed that all deliveries of undiagnosed FMAIT in high-risk women would require caesarean section, preventive therapy would reduce the risk by 50%. The cost-effectiveness of the primiparous strategywould then become $98,000 per FMAIT averted. The threshold chosen to define thrombocytopenia (in terms of platelet count) was also subjected to a sensitivity analysis.
Authors' conclusions The authors conclude that screening newborns for neonatal alloimmune thrombocytopenia is more cost-effective than screening primiparous women whilst admitting that more data are needed before a decision on whether to implement such a screening programme is taken.
CRD COMMENTARY - Selection of comparators Two methods of screening were chosen and compared although a control group for each arm of screening would have been of benefit to the study.
Validity of estimate of measure of benefit Power calculations were performed and although the correct number of primiparous women were included in the study, the number of newborns included appears to be below that required by the power calculations. The effectiveness results were not subjected to statistical testing which, in view of the low number of cases detected, would have been necessary. Three separate maternity wards acted as the sites for this study although centralised testing in one laboratory was undertaken which would provide consistency in terms of the test results.
Validity of estimate of costs The costs of the tests and procedures was the charge to the payer, with no details provided as to how this was calculated. Furthermore this study was undertaken in 3 maternity wards but there was no discussion of possible variations in the costs of tests, procedures and treatments between these wards.
Other issues This study into the effectiveness of screening for MAIT/FMAIT suffers from the fundamental problem that not enough appears to be known about the in-utero treatment of this condition when it is found among pregnant women. For a screening programme to be effective, it is essential that there is an acceptable and effective treatment of the condition being screened for. This does not appear to be the case for the condition under consideration in this study, at least for screening of pregnant women: the test for detecting whether a foetus is affected is a dangerous procedure and one which 50% of the women in this study chose to decline.
Secondly, in the two cases where the condition was detected, different treatment modalities were employed with very different outcomes. With the small numbers under study (2 pregnant women were treated) and no statistical testing, it is very difficult to assess the effectiveness of the treatment modalities and whether screening is actually beneficial. The authors themselves noted in the introduction that "affected foetuses and newborns may recover simultaneously or require therapy" and until more data on this issue are available, it is difficult to assess the benefits of screening. Turning to the newborn strategy, the authors weighted this approach to take account of the fact that information provided will only help future pregnancies and not all women go on to have a second child.
Source of funding Funded in part by a grant from the Fondation de France, Paris (Grant No 900949), and by grants from Delegation Recherche Clinique (Grant Nos 900155 and 912402), Assistance Publique-Hopitaux de Paris, Paris, France.
Bibliographic details Durand-Zaleski I, Schlegel N, Blum Boisgard C, Uzan S, Dreyfus M, Kaplan C, Aujard Y, Baumann C, Blot P, Boissinot C, Daffos F, Fernandez H, Forestier F, Hurtaud-Roux F, Oury J F, Pons J C, Tchernia G, Verdy E. Screening primiparous women and newborns for fetal/neonatal alloimmune thrombocytopenia: a prospective comparison of effectiveness and costs. American Journal of Perinatology 1996; 13(7): 423-431 Indexing Status Subject indexing assigned by NLM MeSH Antigens, Human Platelet /immunology; Cost-Benefit Analysis; Costs and Cost Analysis; Epitopes /immunology; Female; Fetal Diseases /diagnosis /prevention & Humans; Infant, Newborn; Mass Screening /economics; Neonatal Screening /economics; Parity; Pregnancy; Prospective Studies; Thrombocytopenia /diagnosis /prevention & control; control AccessionNumber 21997000037 Date bibliographic record published 31/01/1999 Date abstract record published 31/01/1999 |
|
|
|