Skip to main content
. Author manuscript; available in PMC: 2015 Nov 4.
Published in final edited form as: Lancet Infect Dis. 2015 Mar 16;15(6):692–702. doi: 10.1016/S1473-3099(15)70024-1

Table 4. Risk factors for PCR-confirmed recrudescence at day 28.

Total N (n)* Univariable analyses
Multivariable analyses
PAR calculations
Crude HR (95% CI) p value Adjusted HR (95% CI) p value Frequency§ PAR
Age, every 1-year increase 14 139 (304) 0·96 (0·94–0·99) 0·001 · · · · · · · ·
Bodyweight, every 1 kg increase 14 139 (304) 0·98 (0·97–0·99) 0·001 · · · · · · · ·
Lumefantrine dose, every 5 mg/kg
increase
14 139 (304) 0·98 (0·95–1·02) 0·380 0·98 (0·94–1·02) 0·380 27·46% 6·30%
Enrolment clinical variables
 Parasitaemia, parasites per μL every
ten-times increase
14 139 (304) 1·47 (1·2–1·81) 0·0002 1·41 (1·15–1·74) 0·0012 9·51% 4·01%
 Fever, temperature >37·5°C 13 024 (290) 1·09 (0·85–1·41) 0·500 · · · · · · · ·
 Haemoglobin, every 10 g/L increase 10 303 (221) 0·95 (0·88–1·01) 0·100 · · · · · · · ·
 Anaemia, haemoglobin <100 g/L 10 303 (221) 1·24 (0·93–1·65) 0·150 · · · · · · · ·
 Gametocytes present 9008 (198) 1·01 (0·62–1·65) 0·970 · · · · · · · ·
Sex
 Female 6448 (143) Reference · · · · · · · · · ·
 Male 7393 (156) 0·95 (0·76–1·20) 0·690 · · · · · · · ·
Age category, years
 ≥12 3160 (34) Reference · · · · · · · · · ·
 <1 809 (21) 1·73 (0·92–3·29) 0·091 1·78 (0·89–3·55) 0·100 5·72% 4·76%
 1 to <5 7231 (204) 2·17 (1·35–3·47) 0·001 2·00 (1·23–3·23) 0·005 51·14% 37·92%
 5 to <12 2939 (45) 1·33 (0·81–2·18) 0·260 1·27 (0·76–2·12) 0·360 20·79% 7·96%
Transmission setting
 Low 3432 (60) Reference · · · · · · · · · ·
 High 5336 (160) 1·75 (0·92–3·33) 0·086 · · · · · · · ·
 Moderate 5371 (84) 1·13 (0·58–2·23) 0·720 · · · · · · · ·
Region
 Africa 11 674 (260) Reference · · · · · · · · · ·
 Asia 2306 (43) 0·81 (0·34–1·90) 0·630 · · · · · · · ·
 South America 159 (1) 0·25 (0·02–4·16) 0·330 · · · · · · · ·
Treatment supervision
 Full 10 929 (232) Reference · · · · · · · · · ·
 Partial 1909 (51) 1·40 (0·76–2·59) 0·280 · · · · · · · ·
 Unsupervised 1301 (21) 1·31 (0·52–3·28) 0·570 · · · · · · · ·
Coadministration with fat
 With fatty meal 6346 (142) Reference · · · · · · · · · ·
 Without fatty meal 1181 (23) 0·95 (0·34–2·67) 0·920 · · · · · · · ·
 Advised with fatty meal 2120 (39) 0·98 (0·40–2·41) 0·960 · · · · · · · ·
Dose calculation method
 Per protocol 9473 (231) Reference · · · · · · · · · ·
 Tablet counts 4666 (73) 0·73 (0·43–1·24) 0·250 · · · · · · · ·
Drug tradename
 Coartem 13 891 (296) Reference · · · · · · · · · ·
 Generic artemether–lumefantrine 248 (8) 1·12 (0·42–2·99) 0·820 · · · · · · · ·

HR=hazard ratio. PAR=population-attributable risk.

*

Number of patients (number with recrudescence by day 28).

The assumption of proportional hazard held for the model (p=0·47 for global test) and for all the individual covariates in the multivariable model (p>0·05). The variance of random eff ect was 0·95. The likelihood ratio test was not significant for bodyweight (p=0·19), haemoglobin (p=0·29), and transmission (p=0·53) in the presence of mg/kg dose, parasitaemia, and age category; thus, these were dropped from the multivariable analysis.

Overall PAR for model: 51·1%. Cumulative PAR for parasitaemia >100 000 parasites per μL and age 1 to <5 years: 40·4%.

§

The proportion of patients with the risk factor. Continuous covariates were categorised as follows: baseline parasitaemia at 100 000 parasites per μL and mg/kg lumefantrine dose at 60 mg/kg (lower bound of WHO therapeutic range). The adjusted HR used for estimating the PARs (obtained from the categorised model) were 1·44 for baseline parasitaemia, 1·24 for lumefantrine dose <60 mg/kg, 1·87 for age <1 year, 2·19 for age 1–5 years, and 1·42 for age 5–12 years.