Biosocial determinants of hysterectomy in New Zealand

Objectives. This study examined the prevalence and biosocial correlates of hysterectomy. Methods. Data were from a 1995 national survey of women aged 20 to 59 years . We applied piecewise nonparametric exponential hazards models to a subsam ple aged 25 to 59 to estimate the effects of biosocial correlates on hyster ectomy likelihood. Results. Risks of hysterectomy for 1991 through 1995 were lower than those before 1981. University-educated and professional women were less likely to undergo hysterectomy. Higher parity and intrauterine device side effects i ncreased the risk. Conclusions. This study confirms international results, especially those on education and occupation, but also points to ethnicity's mediating role. E ducation and occupation covary independently with hysterectomy. Analysis of time variance and periodicity showed declines in likelihood from 1981.