The use of socioeconomic deprivation as an explanation of variation in IVF success rates between clinics

Abstract Study question Can differences in the live birth rate (LBR) between clinics with similar standard operating procedures (SOPs) be explained by variations in socioeconomic deprivation? Summary answer Socioeconomic deprivation does not appear to explain differences in LBR between clinics with similar SOPs in this cohort What is known already IVF success rates vary markedly between clinics. Numerous factors determine whether an IVF cycle is likely to be successful. It has been suggested that more socioeconomically deprived women are significantly less likely to achieve a live birth per cycle than less deprived women. In the UK, there is considerable variation in the levels of socioeconomic deprivation between different cities. Study design, size, duration We undertook a multi-center, retrospective review of prospectively collected data. We included all women undergoing their first IVF/ICSI cycle between January 2016 and December 2020 in eight different clinics belonging to the same IVF group. Clinics were located across England: three were situated in ‘very deprived’ cities, two in ‘predominantly affluent’ cities and the remaining cities were ‘neither very deprived nor very affluent’ i.e. their distribution was not skewed. Participants/materials, setting, methods We included all women undergoing their first IVF/ICSI cycle involving transfer of a single, fresh embryo. SOPs for medical and laboratory aspects of treatment were similar across all eight clinics. Socioeconomic deprivation was assessed using the Index of Multiple Deprivation (IMD) determined by the residential postcode of each woman undergoing treatment. Patients were categorized into quintiles according to IMD score. Live birth rates across IMD quintiles were calculated. Statistical analysis was performed using SPSS 22. Main results and the role of chance In total, across the 8 clinics, 17171 women underwent an IVF/ICSI cycle. The LBR ranged from 17.9 - 35.7% (p < 0.001) with both upper and lower extremes in clinics in cities classified as ‘neither very deprived nor very affluent’. The overall LBR ranged from 23.7-26.0% across the IMD quintiles. This difference was neither linear nor statistically significant. Similar trends were noted whether the treatment was NHS (n = 4992) or self-funded (n = 12179). The proportion of women in each IMD quintile varied significantly (p < 0.001) across the different clinics: in the most deprived quintile this ranged from 3.2 - 22.8% and in the least deprived quintile this ranged from 19.1 - 30.8%. The LBR in each IMD quintile also varied significantly (p < 0.001) across the different clinics. The LBR in the most deprived quintile ranged from 7.1 - 46.1% and in the least deprived quintile ranged from 19.4 - 22.9%. Limitations, reasons for caution Our results require caution as the distribution of deprivation amongst women undergoing IVF was not representative of the cities in which clinics were located with a skew towards greater affluence seen. Further unknown variables appear responsible for the significant variation in LBRs between clinics rather than IMD categorization alone. Wider implications of the findings The distribution of deprivation amongst women undergoing IVF was not representative of the cities in which clinics were located. Whilst not unexpected due to the cost of IVF, ethical questions regarding access to fertility care are highlighted and efforts to promote inclusion should be prioritized. Trial registration number Not applicable