What impact did financial incentives for long acting reversible contraceptive (LARC) advice to women have on abortion rates?

Problem

National Institute for Health and Care Excellence (NICE) guidelines recommend long acting reversible contraception (LARC) including injections and implants because they are more cost-effective than other methods such as contraceptive pills to reduce unplanned pregnancies. General practitioners (GPs) were offered financial incentives from 2009 to achieve targets for giving advice about LARC to women aged 13 to 54 who were attending for contraceptive services as part of the quality and outcomes framework (QOF). Although previous research showed an increase in LARC prescriptions after the QOF indicator was introduced in 2009, the policy impact on contraceptive outcomes such as abortion rates is unknown.Aim: To examine trends in recorded abortion as a proxy marker for adverse contraceptive outcome before and after QOF indicator was introduced in 2009 in UK general practice.

Approach

Design: interrupted time series (ITS) to compare trends in abortion rates before and after 2009. We used the Clinical Practice Research Datalink (CRPD) - a database of electronic records from participating practices. We summed clinical procedure and referral codes for termination of pregnancy and abortion, recorded between 2004 and 2014, in women between the ages of 13 and 54.We calculated annual denominator populations of women aged at least 13 years old and no older than 54 that year, registered with a practice which was deemed “up to standard” for data quality the year before. We used crude annual abortion rates per 1000 women for ITS analysis. We conducted a sensitivity analysis with and without a one-year “phase-in” period - time needed for clinicians to implement the scheme and for women to respond to LARC advice. We tested our final models for auto-correlation.

Findings

There were 166,213 abortions recorded on CPRD among 3,281,677 women between ages of 13 and 54 during our study period. Abortion rates fell from 17.2 per 1000 women in 2004 to 7.8 per 1000 women in 2014.Without a phase-in period, there was an annual decline in abortion rates of 0.33 per 1000 women from 2004 to 2008, (95%CI -0.40 to -0.27, p<001), a modest step change in 2009 (-0.47, -0.72 to -0.23, p=0.007), and a steeper post-intervention decline of 0.93 (-0.99 to -0.86, p<0.001). Analysis with phase-in period resulted in a similar annual decline of 0.32 before intervention (-0.37 to -0.27, p<0.001), but a larger step change (-1.37, -1.56 to -1.18, p<0.001), and steeper post-intervention decline of 0.97 (-1.05 to -0.89, p<0.001).

Consequences

Abortion rates declined sharply after introduction of the QOF indicator to offer LARC advice in 2009. By 2014, five years after the incentives, abortion rates were 47% lower than would have been expected without the incentive.