BACKGROUND: Aim of this study is to define the risk factors of
ovarian hyperstimulation syndrome onset during assisted
reproduction treatment using in vitro fertilization technique.
Exploratory computer analysis of electronically stored data
about assisted reproduction treatment cycles in clinical
registry with the use of data mining system. MATERIALS:
Analyzed file included data of 12 527 monitored cycles from
1989 to 2003. Cycles which leaded to development of ovarian
hyperstimulation syndrom were analyzed (2456 cases, 19,6 % of
cycles). METHODS: Both the ovarian hyperstimulation syndrome
complicated cases and cases without ovarian hyperstimulation
were tested by data mining method which is designed to find
statistically significant differences among input attributes of
ovarian stimulation phase of therapeutic cycles. The observed
differences between input attributes were statistically tested
and the value of statistical significance was evaluated.
RESULTS: Significantly higher incidence of a clinically
important form of ovarian hyperstimulation syndrome development
was observed among patients under 30 years old who were
affected by hyperstimulation syndrome in previous treatment
cycles. We can predict a higher risk in patients with
oligomenorrhoea and in cases of immunological and andrological
sterility factor. Higher incidence of clinically significant
form of hyperstimulation syndrome is noticed with stimulation
protocols with highly effective gonadotrophine hormones where
the total dose of gonadotrophins is greater than 1125 IU and
maximum 17-beta estradiol levels are greater than 20 nmol/l.
CONCLUSIONS: We have proved the applicability of the data
mining system for the analysis of the risk factors with
influence on the Assisted reproduction treatment results. Data
mining method can be used to define statistically significant
relations among all attributes of ovarian stimulation part of
therapeutic cycle and to define the risk factors of
hyperstimulation syndrome occurrence. Key words: ovarian
hyperstimulation syndrom, risk factors, sterility treatment,
data mining.