Cost of fertility treatment and live birth outcome in women of different ages and BMI

Study question: What is the impact of different age and BMI groups on total investigation and treatment costs in women attending a secondary/tertiary care fertility clinic?

Summary answer: Women in their early to mid-30s and women with normal BMI had higher cumulative investigation and treatment costs, but also higher probability of live birth.

What is known already: Female age and BMI have been used as criteria for rationing publically funded fertility treatments. Population-based data on the costs of investigating and treating infertility are lacking.

Study design, size and duration: A retrospective cohort study of 2463 women was conducted in a single secondary/tertiary care fertility clinic in Aberdeen, Scotland from 1998 to 2008.

Participants/materials, setting, methods: Participants included all women living in a defined geographical area referred from primary care to a specialized fertility clinic over an 11-year period. Women were followed up for 5 years or until live birth if this occurred sooner. Mean discounted cumulative National Health Service costs (expressed in 2010/2011 GBP) of fertility investigations, treatments (including all types of assisted reproduction), and pregnancy (including delivery episode) and neonatal admissions were calculated and summarized by age (≤ 30, 31-35, 36-40, >40 years) and BMI groupings (<18.50, 18.50-24.99 (normal BMI), 25.00-29.99, 30.00-34.99, ≥ 35.00 kg/m(2)). Further multivariate modelling was carried out to estimate the impact of age and BMI on investigation and treatment costs and live birth outcome, adjusting for covariates predictive of the treatment pathway and live birth.

Main results and the role of chance: Of the 2463 women referred, 1258 (51.1%) had a live birth within 5 years, with 694 (55.1%) of these being natural conceptions. The live birth rate was highest among women in the youngest age group (64.3%), and lowest in those aged >40 years (13.4%). Overall live birth rates were generally lower in women with BMI >30 kg/m(2). The total costs of investigations were generally highest among women younger than 30 years (£491 in those with normal BMI), whilst treatment costs tended to be higher in 31-35 year olds (£1,840 in those with normal BMI). Multivariate modelling predicted a cost increase associated with treatment which was highest among women in the lowest BMI group (across all ages), and also highest among women aged 31-35 years. The increase in the predicted probability of live birth with exposure to treatment was consistent across age and BMI categories (∼ 10%), except in the oldest age group where a slightly smaller increase in the probability of live birth was observed. The ratio of increased costs to the increased probability of live birth in women who were treated increased markedly in women over the age of 40 years, but tended to fall as BMI increased within all age groups.

Limitations and reason for caution: Our results, based on retrospective observational data from a single centre, have limited generalizability and are not free from clinician and clinic selection bias which can influence the choice of treatments as well as their costs.

Wider implications of the findings: Spontaneous live birth rates were particularly high in younger women with unexplained infertility, suggesting that expectant management is a reasonable option in this group. The policy of not over-investigating older women and offering early treatment where appropriate still incurred the highest costs per additional live birth associated with treatment, owing to the lower probability of treatment success. The increased additional cost for each live birth associated with treatment for women with decreasing BMI across all age groups, suggests that it may be possible to identify a more targeted approach to treatment.

Study funding/competing interests: This study was partly funded by an NHS endowment grant (Grant Number 12/48) and D.J.M. by a Chief Scientist Office Postdoctoral Fellowship (Ref PDF/12/06). There are no conflicts of interest to declare.