1 Centre for Epidemiology, National Board of Health and Welfare, SE-106 30, 2 Sophiahemmet, SE-114 86, 3 Centre for Epidemiology, National Board of Health and Welfare, SE-106 30 Stockholm and 4 Tornblad Institute, University of Lund, SE-232 65 Lund, Sweden
![]() |
Abstract |
---|
![]() ![]() ![]() ![]() ![]() |
---|
Key words: cerebral palsy/costs/epilepsy/hospitalization/IVF
![]() |
Introduction |
---|
![]() ![]() ![]() ![]() ![]() |
---|
We undertook a study of IVF infants using central health registers in Sweden in order to study frequency and amount of hospitalization among these infants compared with all infants born in Sweden.
![]() |
Material and methods |
---|
![]() ![]() ![]() ![]() ![]() |
---|
The hospitalizations of each infant in the Medical Birth Register were ascertained by record linkage with the Hospital Discharge Register. This register contains information on all hospitalizations in Sweden, including diagnoses, operations performed and dates of entrance to, and discharge from, the hospital. Each person is identified by their personal identification number, which enables linkage with other registers. The register covers the whole country from 1987 onwards. For the years 19841986 a few counties were missing, but only 69 IVF infants were born during these years and information loss is small and probably similar for IVF births and all births. The registry contains information on discharge events for 557 990 infants.
We studied the proportion of infants born after IVF who had been hospitalized at any age (up to a maximum of 14 years) and for various age windows. This analysis was made irrespective of the number of infants in the birth and also comparing singleton IVF infants with singleton controls and IVF twins with control twins. Analysis was made using the MantelHaenszel estimator procedure with stratification for year of birth, maternal age (5-year classes), parity (14+), smoking habits in early pregnancy (unknown, none, <10 and 10 cigarettes per day), and years of involuntary childlessness (05+ years). A separate analysis was made on term infants (gestational duration
37 weeks or if gestational duration was unknown, a birth weight of
2500 g).
We then repeated the analyses restricted to a number of specified diagnoses: cerebral palsy [international classification of diseases (ICD)-9 343, ICD-10 G80], epilepsy (ICD-9 345, ICD-10 G40 or G41), mental retardation (ICD-9 317319, ICD-10 F70-F79), attention deficit hyperactivity disorder (ICD-9 314, ICD-10 F90), non-psychotic disorders of psychological development (ICD-9 315, ICD-10 F80-F83) or autism and related conditions (ICD-9 299, ICD-10 F84), accidents (ICD-9 800995, ICD-10 S00-T98), tumours (ICD-9 140239, ICD-10 C00-D48), asthma (ICD-9 493, ICD-10 J45 or J46) after the age of 1 year, any infection, or any congenital malformation (Chapter 14 in ICD-9, Chapter 17 in ICD-10). Most of these diagnoses occur at an excess rate in children which are born preterm.
We calculated the mean number of days spent in hospital. If date of entrance and discharge were the same, duration was called 1 day. A difference was formed between the average number of days spent in hospital by IVF children and that spent by all other children.
By record linkage with the Swedish Cancer Register, we identified children who developed childhood cancer before the end of 1998. The odds ratio (OR) for developing cancer was determined, stratifying for year of birth, maternal age, parity and known period of involuntary childlessness. No stratification for smoking was made as it has been shown that smoking is a no risk factor for childhood cancer (Pershagen et al., 1992).
Results
Figure 1 shows some maternal characteristics which affect the risk for a child to be hospitalized. There is an excess risk at low maternal age, high parity, smoking and involuntary childlessness. Each factor increases the risk of hospitalization independently of the other factors.
|
|
|
|
Table III shows the average number of days spent in hospitals (calculated for all children born) in IVF and non-IVF children. It can be seen that IVF children have, on average, spent 6 more days in hospital than non-IVF children. When singleton IVF children are compared with singleton non-IVF children, an excess period of hospitalization is still seen amounting to 3 days. A similar comparison between IVF twins and non-IVF twins shows no excess hospitalization, but non-IVF twins were hospitalized 23 times longer than non-IVF singletons.
|
Among infants born in 19911993 (and thus on average 6.5 years old at the end of follow-up), IVF children spent an average of 10.7 days in hospital, while non-IVF children spent an average of 3.5 days, a difference of 7.2 days. In 1996, ~1500 IVF children were born in Sweden. Their excess hospital care before the age of 6 years can thus be estimated to be ~10 800 days, with an average cost of 5000 SEK (~US$500) per hospital day (Statistisk Årsbok för Landsting, 1999); this means an excess cost of 54x106 SEK.
Discussion
This study confirms and extends previously published observations that the main increased utilization of health care by IVF children occurs during the first period of life (Leslie et al., 1998), even though we found an increased use up to the age of 6 years. The increase is mainly due to the occurrence of twins, but IVF singletons are also more often hospitalized than non-IVF singletons, perhaps partly reflecting the increased frequency of preterm births. When the analysis is restricted to term infants, there is still an apparent increase, but this disappears when the length of involuntary childlessness (when known) is taken into consideration. There may be a parental factor involvedparents of children who were born after a long period of involuntary childlessness may be more concerned over child morbidity and may seek medical advice more easily than other parents.
When the risk for hospitalization is compared between IVF and non-IVF children, the excess risk is higher when singletons rather than twins are studied. The latter excess risk is probably lower because one is comparing IVF twins, which are mainly dizygotic, with non-IVF twins, which in a large proportion are monozygotic (and with a higher disease risk).
The excess risk of hospitalization among IVF children was found for many different conditions. The highest OR was found for congenital malformations followed by cerebral palsy, but there are also statistically significantly increased ORs for conditions like asthma and infections. For such conditions, the possibility of an effect of increased parental concern must be considered, but prematurity may play a direct role. An increased OR for tumours (based on 43 children) was not verified when the Cancer Register was used (only 11 children were identified and no increased risk). The possibility of an increased cancer risk in IVF children has been discussed based on case reports, but has never been proved. A study of 332 IVF children found no cancer case, but the power to detect a risk increase is low (Lerner-Geva et al., 2000). The overuse of hospital care for benign tumours may again be an expression of parental concern. There are, however, four children with histiocytosis against the expected number of 0.35, but this finding could be due to mass significance. Support or dismissal from other studies is needed.
The figures on the number of days of hospitalization can be used for estimating the `extra' cost of paediatric care of IVF children and notably of those born as twins. Most interesting is the impact of twinning after IVF, because this is to a large extent avoidable by the transfer of only one embryo. An IVF twin born in 19911993 used as an average 13 days of hospital care against 5.6 days for a singleton IVF child, a difference of 7.4 days, corresponding to ~37 000 SEK. To this should be added the costs outside the hospital care cost, specifically associated with cerebral palsy and other severe handicaps. These costs can be compared with the extra cost associated with additional transfers because more transfers will be needed in order to achieve a pregnancy. Cost estimates and comparisons differ between different countries. In Sweden, the savings in hospital care cost would possibly pay for one or two extra transfers.
In conclusion, our study shows that IVF children have an increased use of in-patient care, not only in the neonatal period but for some years afterwards. By reducing the rate of twin births by using one-embryo transfers a substantial saving in hospital cost could be achieved, which could balance the extra number of embryo transfers needed in order to obtain a pregnancy.
![]() |
Notes |
---|
![]() |
References |
---|
![]() ![]() ![]() ![]() ![]() |
---|
Cnattingius, S., Ericson, A., Gunnarskog, J. and Källén, B. (1990) A quality study of a medical birth registry. Scand. J. Soc. Med., 18, 143148.[ISI][Medline]
Gissler, M., Silverio, M.M. and Hemminki, E. (1995) In-vitro fertilization pregnancies and perinatal health in Finland 19911993. Hum. Reprod., 10, 18561861.[Abstract]
Lerner-Geva, L., Toren, A., Chetrit, A., Modan, B., Mandel, M., Rechavi, G. and Dor, J. (2000) The risk for cancer among children of women who underwent in vitro fertilization. Cancer, 88, 28452847.[ISI][Medline]
Leslie, G.I., Bowen, J.R., Arnold, J.D. and Saunders, D.M. (1992) In-vitro fertilisation and neonatal ventilator use in a tertiary centre. Med. J. Aust., 157, 165167.[ISI][Medline]
Leslie, G.I., Gibson, F.L., McMahon, C., Tennant, C. and Saunders, D.M. (1998) Infants conceived using in-vitro fertilization do not overutilize health care resources after the neonatal period. Hum. Reprod., 13, 20552059, 1998.[Abstract]
Pershagen, G., Ericson, A. and Otterblad Olausson, P. (1992). Maternal smoking in pregnancy: does it increase the risk of childhood cancer? Int. J. Epidemiol., 21, 15.[Abstract]
Statistisk Årsbok för Landsting (1999) Statistical Year-book for County Councils, Table II:20, Stockholm.
Submitted on August 23, 2001; accepted on December 6, 2001.