Though Makadzange et al. efforts are commendable, several safety, methodologic, and reporting issues deserve clarification [1]. First is safety. Fluconazole increases the area-under-the-curve (AUC) of nevirapine in a dose-dependent manner. Nevirapine AUC increases by 29% with 200mg fluconazole given three times weekly, by 50% with 200mg fluconazole daily, and by 90–100% with 400mg fluconazole daily [2–4]. Effects of fluconazole 800mg daily on nevirapine pharmacokinetics are unknown. As only immediate ART subjects were exposed to 800mg fluconazole and nevirapine for 10 weeks, lack of safety monitoring data is not reassuring that increased hepatotoxicity was not contributory to excess mortality. The authors reported, “Liver function test results remained normal throughout the study …” Though this may technically be true, only 44 subjects had baseline ALT measured and monitoring was reported as occurring every 6 months. When nearly half the deaths (17/35) occurred before first follow-up, 15 of 35 occurred at home, and 33 of 35 occurred prior to six months, an assumption of no hepatotoxicity is incorrect in the absence of actual data. Second, the censoring is unclear. The reported mortality with immediate ART was 88% of 28 persons. No integer divided by 28 equals 88%. Based on the text, reported mortality in immediate ART arm was 82.1% (23/28) vs. delayed arm 46.1% (12/26). Of more concern is that outcomes displayed in Figure 2 do not match mortality (n=12) and lost (n=3) for the delayed arm, nor mortality (n=23) and lost (n=3) for the immediate arm as reported in the text and Figure 1. In counting the ‘number at risk’ in Figure 2, for delayed and immediate ART arms, 15 and 24 respective subjects are dead or lost-to-follow up by 200 days, with one additional death in each arm after 200 days. Thus, non-survival rates are: 61.5% (16/26) for delayed vs. 89.3% (25/28) for immediate ART (P=.026). Based on reported mortality (i.e. 12 and 23 deaths), the Kaplan-Meier figure thereby implies 4 subjects were censored in the delayed arm and 2 in the immediate arm. This censoring would yield 54.5% (12/22) and 88% (23/26) mortality (similar to that reported by the authors). The implication is that either participant numbers or outcomes in Figure 1, Figure 2, and/or the text are incorrect as well as that intent-to-treat analysis was not consistently utilized, though stated as such. One fundamental question for any clinical trial is appropriate outcome measures. We would advocate that survival is a better outcome measure than known mortality. Cryptococcal meningitis has a near 100% mortality within 6 months in the absence of ART [5, 6]. In a time-to-event analysis, right-hand censoring is generally considered acceptable for lost-to-follow up, yet we question the appropriateness of doing so in all scenarios. This is particularly germane to 3 (or possibly 4) subjects in the delayed arm who were lost-to-follow-up before 2 weeks and did not initiate ART. Considering them a success by time-to-event analysis before censoring at ≤2 weeks creates a subtle systematic bias. The overall concern is a type-I error due to relatively small sample size, underpowered nature of the trial, inadequacy of randomization (e.g. CD4’s unequal), post-hoc clinicaltrials.gov registration, and further complicated by early stoppage. Unexplained shifts in 1–2 subjects become potentially important. Nevertheless, the overall conclusion is likely correct that after CM when using nevirapine with fluconazole 800mg, immediate ART results in higher mortality. The authors’ conclusion that “early initiation of ART most likely resulted in increased rates of cryptococcal IRIS” is unsubstantiated speculation, as no attempt was made to assess for cryptococcal-IRIS. The final question is whether this trial’s result is generalizable, particularly if efavirenz were used instead of nevirapine, nevirapine dose adjusted to 200mg daily while receiving fluconazole 800mg, amphotericin used, or degree of immunosuppression fully equal. Further trials are needed.