Comparison of statistical models to estimate parasite growth rate in the induced blood stage malaria model

Malaria Journal

Table 3 Parasite growth rate estimates for models fitted overall by study using mixed effects models (models 7–12)

Model used	Parameter estimated	Sanderson et al.		Duncan et al.		Payne et al.		SD across studies fixed/non-fixed intercept model
Model used	Parameter estimated	Fixed intercept (n = 5)	Non-fixed intercept (n = 5)	Fixed intercept (n = 8)	Non-fixed intercept (n = 8)	Fixed intercept (n = 27)	Non-fixed intercept (n = 27)	SD across studies fixed/non-fixed intercept model
Log-linear	m (SE) [95% CI]	0.63 (0.025) [0.58–0.68]	0.67 (0.084) [0.50–0.84]	0.53 (0.013) [0.50–0.55]	0.69 (0.055) [0.58–0.80]	0.52 (0.010) [0.50–0.54]	0.66 (0.024) [0.61–0.71]	0.061/0.017
Sine-wave (fixed period)	m (SE) [95% CI]	0.63 (0.024) [0.58–0.67]	0.67 (0.083) [0.51–0.83]	0.53 (0.015) [0.50–0.56]	0.74 (0.036) [0.67–0.81]	0.52 (0.010) [0.50–0.54]	0.68 (0.022) [0.63–0.72]	0.059/0.040
Sine-wave (non-fixed period)	m (SE) [95% CI]	0.62 (0.024) [0.58–0.67]	0.71 (0.068) [0.58–0.84]	0.53 (0.016) [0.50–0.56]	0.73 (0.035) [0.66–0.80]	0.67 (0.014) [0.65–0.70]	0.70 (0.015) [0.67–0.72]	0.071/0.018

Fixed-intercept models had the intercept fixed to the inoculum size, whereas non-fixed-intercept models estimated the intercept as part of model fitting. Sine-wave models were fitted either with the period fixed to 48 h (fixed period), or with the period not fixed (non-fixed period). The parasite growth rate estimate (m) overall by study was estimated by the mixed effects model
m parasite growth rate estimate, SD standard deviation, SE standard error, CI confidence interval

ISSN: 1475-2875