The article is also available on the Medical Observer website (may need registration).
Zhongyu, a 50-year-old allied health practitioner presented with a recurrence of podagra – he’d had several attacks of acute gout in past year. As we spoke about preventive therapy, I remembered an online discussion at GPs Down Under – is the recommendation against starting allopurinol in acute gout evidence-based?
Does initiating allopurinol in acute gout delay recovery from the episode?
What does the research evidence say?
Step 1: The Cochrane Library
The Cochrane Library did not have a systematic review that answered this question.
Step 2: TripDatabase
Next, I conducted a search using the TripDatabase PICO search tool (P: “acute gout”, I: “allopurinol”, C: “placebo”, O: blank). The top article returned was a randomised trial on this question by Hill et al. (2015) published in the Journal of Clinical Rheumatology.  Let’s look at this paper in more detail.
I will use the randomised controlled trial appraisal sheet from the Centre for Evidence Based Medicine. 
Participants: who was studied?
37 adult patients (mean age: 57 years, 94% male), with crystal-proven acute gout by arthrocentesis, within 72 hours of initial therapy, referred to the study site at San Antonio Military Medical Centre (Texas). A third had tophi, a quarter erosions. About 90% had received corticosteroids (either intramuscular, intra-articular, or oral) for acute treatment. “Prophylaxis” therapy was with colchicine or meloxicam.
Important exclusions: GFR < 50 mL/min, LFT abnormality, women of childbearing age, use of allopurinol in the past 6 months.
Intervention: what was the exposure?
allopurinol 100 mg daily × 14 days, then 200 mg daily × 14 days
Comparator: what was the control/alternative?
placebo capsules, given in the same regimen
Outcomes: what was measured?
Primary outcome: days to resolution of acute gout. “Resolution” was defined by a combination of physical examination findings and patient-rated symptoms.
Internal validity: are the trial results valid?
Randomised patient assignment?
Yes. This appeared well performed and described (p. 121). 
Groups similar at the start?
No. The groups had some notable differences. The allopurinol group was on average older, and with fewer previous attacks (Table 1, p. 122). 
Groups treated equally apart from assigned treatment?
No. The allopurinol group received more intramuscular steroids (62.5% vs 42.1%) and less intra-articular steroids (12.5% vs 42.1%).
All patients accounted for?
Yes. Few participants dropped out of the study.
Measures objective? Or patients and clinicians kept blinded?
Unclear. Although the study was double-blinded, physicians guessed the randomisation correctly 84% of the time (p. 123). 
What were the results?
From the intention-to-treat analysis, the average number of days to resolution from gout:
- placebo group: 12.53 days (standard deviation 7.73)
- allopurinol group: 17 days (SD 8.53)
- mean difference: 4.47 days (P = 0.13)
Unfortunately, this study has potential critical threats to internal validity. Although the participants were randomised, there were important differences between the two groups. The blinding may have been compromised with the majority of physicians correctly guessing the group allocation. The consequence is that we cannot be confident that the results found are due to allopurinol versus placebo, rather than other causes, e.g., differences in age of the participants, other treatment they received, and other unknown biases.
The assertive conclusion by the study authors that “in this cohort, allopurinol did not prolong the acute, treated attack”, despite the point estimate favouring placebo by 4.5 days, is problematic. The authors seem to have overvalued the importance of the non-significant P-value, using it to conclude that the null hypothesis (that there was no difference between groups) was true. This interpretation is invalid, especially in the setting of an underpowered study where false negatives are common (see Stat Facts). 
What else has been published on initiating allopurinol in acute attacks? The only other study was small and found no significant differences.  eTG Complete recommends against it,  the American College of Rheumatology supports the practice,  and a recent Australian clinical update is silent on the matter. 
Effectively, there is no clear evidence of the effect that allopurinol initiation during an acute attack has on symptoms. In practice, it may be preferable to avoid starting allopurinol in an acute attack, but this should not be a dogmatic stance. In certain local contexts and situations, early commencement of allopurinol may be a reasonable pragmatic option.
Statistical power, and low powered studies
Low powered studies reduce the chance of detecting a true effect.  My post hoc calculation of the results suggests that the experiment’s statistical power was less than 30%. Moreover, the study design had less than 10% power to detect a small effect-size. That is, if allopurinol has a small to moderate effect on the time to recovery in acute gout, this study had no more than a 10-30% chance of detecting it.
- Hill EM, Sky K, Sit M, Collamer A, Higgs J. Does starting allopurinol prolong acute treated gout? A randomized clinical trial. J Clin Rheumatol 2015;21(3):120-5. DOI: 10.1097/RHU.0000000000000235
- RCT Appraisal Sheets [website]. Centre for Evidence Based Medicine. Retrieved on: 2015 October 15. Available from: http://www.cebm.net/critical-appraisal/
- Button KS, Ioannidis JP, Mokrysz C, Nosek BA, Flint J, Robinson ES, et al. Power failure: why small sample size undermines the reliability of neuroscience. Nat Rev Neurosci 2013;14(5):365-76. DOI: 10.1038/nrn3475
- Taylor TH, Mecchella JN, Larson RJ, Kerin KD, Mackenzie TA. Initiation of allopurinol at first medical contact for acute attacks of gout: a randomized clinical trial. Am J Med 2012;125(11):1126-34 e7. DOI: 10.1016/j.amjmed.2012.05.025
- eTG Complete (etg45 July 2015). Revised on: 2010 Oct. Retrieved on: 2015 Oct 15. Available from: http://etg.tg.com.au/ip/desktop/tgc/rhg/2037.htm
- Khanna D, Fitzgerald JD, Khanna PP, Bae S, Singh MK, Neogi T, et al. 2012 American College of Rheumatology guidelines for management of gout. Part 1: systematic nonpharmacologic and pharmacologic therapeutic approaches to hyperuricemia. Arthritis Care Res 2012;64(10):1431-46. DOI: 10.1002/acr.21772
- Ting K, Graf SW, Whittle SL. Update on the diagnosis and management of gout. MJA 2015;203(2):86-8. DOI: 10.5694/mja14.00953