Systematic reviews are often interested in the change in health outcomes over time, aggregated over multiple timepoints and studies. Meta-regression may be used to estimate such trajectories. However, it is often the case that included studies present different outcome assessment patterns over time. It is not known how well existing meta-regression methods perform in the presence of such heterogeneity. Simulated data from an individual participant data meta-analysis will be used to compare two existing methods to two proposed extensions. White et al. (2019) proposed using fractional polynomial (FP) basis expansions within linear mixed models to meta-estimate trajectories. We propose two extensions which use spline rather than FP expansions. Results from a simulation that considers varying the level of heterogeneity in the timing and number of assessments between studies will be presented.