%0 Figure %A Tinh Hien, Tran %A F. Boni, Maciej %A E. Bryant, Juliet %A Thuy Ngan, Tran %A Wolbers, Marcel %A Dang Nguyen, Tran %A Thanh Truong, Nguyen %A Thi Dung, Nguyen %A Quang Ha, Do %A Minh Hien, Vo %A Tan Thanh, Tran %A Nguyen Truc Nhu, Le %A Thi Tam Uyen, Le %A Thi Nhien, Pham %A Tran Chinh, Nguyen %A Van Vinh Chau, Nguyen %A Farrar, Jeremy %A Rogier van Doorn, H. %D 2010 %T Per patient analysis of RT-PCR results shown by day of illness and day of treatment. %U https://plos.figshare.com/articles/figure/_Per_patient_analysis_of_RT_PCR_results_shown_by_day_of_illness_and_day_of_treatment_/519957 %R 10.1371/journal.pmed.1000277.g004 %2 https://plos.figshare.com/ndownloader/files/849520 %K rt-pcr %K shown %X

Time to PCR negativity and its dependence on illness-to-treatment interval. (A) Gray lines show the minimum and maximum number of patients who were PCR-negative after a certain number of days of illness, on the basis of patients' last positive PCR result and first negative PCR result, which could be separated by a gap of as many as 4 d. Red line shows the ML-fit (see Methods) of time to PCR negativity, and dashed lines are 95% confidence bands. (B) as (A), related to days of treatment. (C, D) Minimum and maximum durations of PCR positivity for patient subgroups corresponding to the length of illness-to-treatment interval. (E, F) ML-curves describing time to PCR-negativity for patient subgroups. Curves for patients who started treatment on the day of illness onset (illness-to-treatment interval = 0, 11 patients), and patients who started treatment 5 d postillness (illness-to-treatment interval = 5, 10 patients) are shown in gray as they differ qualitatively from the other four curves. Legend in (D) applies to (C–F). Data from 278 patients with both negative and positive PCRs were used to make these graphs.

%I PLOS Medicine