Results
Under the exclusion criteria specified for the PTs ‘Angioedema’ and ‘Hepatitis’ (see Table S1 for studied LLTs), 36,654 ICSRs with single suspected drugs and removed suspected duplicates were retrieved for this study. The proportion of excluded reports with imprecise (i.e. partly missing) or potentially imprecise (i.e. start or onset day 01 and/or 15) RTTO was 41% for hepatitis and 24% for angioedema. Out of these, the potentially imprecise dates accounted for 25% for hepatitis compared to 16% for angioedema. Other applied exclusion criteria for assuring data quality (specified in section 2.5) only affected 4% of the data on hepatitis versus 6% on angioedema.
The analysis of RTTO for the selected short- and long-term treatments included 4,139 reports, the demographics of which are presented in Table S2. As groups, drugs with expected long-term treatments presented significantly longer RTTO than drugs with expected short-term treatments for both angioedema and hepatitis, (with significance levels << 0.001 for both ADRs). In fact, the patterns of RTTO of angioedema for long-term treatments were very similar to the patterns of RTTO of hepatitis for short-term treatments, see Figure 1. Median RTTO for angioedema with long-term treatments ranged 0-26 days, with a median value at 8 days. Fluoxetine presented the highest median RTTO for angioedema among the studied drugs with a latency of 26 days. The corresponding range of median RTTO for hepatitis with short-term treatments was 4-11 days, with a median value at 7.5 days, see Figure 2.
10.1371/journal.pone.0068938.g001 Figure 1  Reported time-to-onset for short-versus long-term treatments for angioedema and hepatitis, respectively.
Drugs with expected long-term treatments (panels B and D) generally presented with longer RTTO than drugs with expected short-term treatment (panels A and C). In fact, the patterns of RTTO of angioedema for long-term treatments were very similar to the patterns of RTTO of hepatitis for short-term treatments (panels B and C).
10.1371/journal.pone.0068938.g002 Figure 2  Median RTTO for short-versus long-term treatments for angioedema and hepatitis, respectively.
The plot visualises the overlap of RTTO medians of short-term treatments for hepatitis with those of long-term treatments for angioedema. The median value of medians for RTTO of angioedema with long-term treatments was 8 days and the corresponding median for hepatitis with short-term treatments was 7.5 days.   In order to investigate the effect of excluding potentially imprecise dates (i.e. start or onset day 01 and/or 15) from the dataset, a sensitivity analysis including the reports with potentially imprecise RTTO for angioedema and hepatitis with the twelve evaluated drugs was conducted. However, it demonstrated that the effect on outcome RTTO was negligible (results not presented here).