Evidence from two systematic reviews supports the use of standard nonsteroidal anti-inflammatory drugs (NSAIDs) for the management of primary dysmenorrhoea. There is insufficient evidence to indicate whether any one NSAID is more effective than another for the treatment of dysmenorrhoea. Low quality evidence suggests that the efficacy of cyclo-oxygenase-2 (COX-2) inhibitors is superior to placebo, and similar to that of NSAIDs. The COX-2 inhibitors rofecoxib and valdecoxib have since been withdrawn from the UK market because of cardiovascular concerns and potentially life-threatening skin reactions.

Standard nonsteroidal anti-inflammatory drugs (NSAIDs)

CKS found one Cochrane review of standard NSAIDs for primary dysmenorrhoea (search date: to May 2009) [Marjoribanks et al, 2010].

• Data from 32 randomized controlled trials (RCTs) comparing NSAIDs with placebo were pooled. Pooled analysis found that NSAIDs were significantly better than placebo at producing moderate or excellent pain relief (odds ratio [OR] 4.50, 95% CI 3.85 to 5.27). However, there was substantial heterogeneity for this finding.
• Two studies compared ibuprofen with paracetamol and one compared naproxen with paracetamol. Pooled data from these three studies found more women using NSAIDs reported good, excellent, or complete pain relief compared with women using paracetamol (OR 1.89, 95% CI 1.05 to 3.43).
• There were 15 RCTs comparing one NSAID with another NSAID. Most of the comparisons were based on small trials, which were unsuitable for meta-analysis. However, the authors concluded that there was little evidence of the superiority of any individual NSAID with regard to either efficacy or safety.

CKS found an earlier systematic review of minor analgesics for primary dysmenorrhoea (including aspirin, naproxen, ibuprofen, and mefenamic acid; search date: to March 1997) [Zhang and Li Wan Po, 1998].

• The systematic review identified 23 RCTs comparing naproxen with placebo, 17 RCTs comparing ibuprofen with placebo, and five RCTs comparing mefenamic acid with placebo.
• The pooled response rate ratios for pain relief were:
• In favour of naproxen, 3.17 (95% CI 2.72 to 3.65).
• In favour of ibuprofen, 2.41 (95% CI 1.58 to 3.68).
• In favour of mefenamic acid, 2.03 (95% CI 1.65 to 2.48).
• Three women would have to be treated with naproxen (NNT 3) for one woman to experience pain relief, and two women would have to be treated with ibuprofen (NNT 2) or mefenamic acid (NNT 2) for one women to experience pain relief.
• The systematic review identified eight RCTs (n = 486) comparing aspirin (usual dose 650 mg, four times a day) with placebo. Pooled data from the eight RCTs showed that:
• Compared with placebo, aspirin significantly increased the proportion of women who experienced at least moderate pain relief (RR 1.60, 95% CI 1.12 to 2.29). Ten women would need to be treated with aspirin for one woman to experience at least moderate pain relief (NNT 10, 95% CI 5 to 50).
• There was a high drop-out rate in the studies of aspirin (27% overall), which questions the reliability of the results.
• There was no evidence of a significant difference in the incidence of adverse effects between the two groups (7–17% with aspirin compared with 3–17% with placebo; RR 1.3, 95% CI 0.79 to 2.17).
• The systematic review included 12 RCTs comparing one NSAID with another NSAID. Most of the comparisons involved small single trials, and it is difficult to draw conclusions regarding comparative efficacy.
• Pooled data from three RCTs did not show a significant difference in pain relief between naproxen and ibuprofen (RR 1.08, 95% CI 0.79 to 1.48).

Cyclo-oxygenase-2 (COX-2) inhibitors

CKS found one meta-analysis of single-dose rofecoxib for primary dysmenorrhoea [Edwards et al, 2004].

• The meta-analysis used data from three industry-sponsored crossover RCTs comparing rofecoxib 25 mg or 50 mg with placebo or an active comparator; the comparator was ibuprofen 400 mg in one study and in naproxen 550 mg in two studies. Active or placebo drugs were randomly assigned over three consecutive menstrual cycles. Data from 231 women were included in the analysis.
• Pain intensity was measured using a 4-point pain intensity scale (0, none; 1, mild; 2, moderate; 3, severe) and pain relief was measured using a 5-point pain relief scale (0, none; 1, a little; 2, some; 3, a lot; 4, complete). The time at which a woman requested a second dose of trial medication (or rescue analgesic), if required, was also recorded.
• All active treatments were significantly more effective than placebo at 6, 8, and 12 hours. For all the active treatments the proportion of women with at least 50% pain relief was about 60% at all time points, and with placebo it was about 30% at all time points.
• There were no significant differences in efficacy between the active treatments. After 12 hours the proportion of women who required additional medication was 28% of those receiving rofecoxib 25 mg, 27% receiving rofecoxib 50 mg, 29% receiving naproxen 550 mg, 41% receiving ibuprofen 400 mg, and 50% receiving placebo. Confidence intervals were not presented.
• Few adverse effects were reported, and none were considered to be serious in any of the trials.

CKS found a technology appraisal (search date: to April 2005) of COX-2 inhibitors for the treatment of dysmenorrhoea [Alberta Heritage Foundation for Medical Research, 2005].

• The technology appraisal identified three published studies that reported results from four RCTs. In one RCT 337 women were randomly assigned to receive meloxicam 7.5 mg, meloxicam 15 mg, or mefenamic acid 500 mg, and were followed up for three cycles. The other two studies were crossover studies in which a total of 256 women received both trial and comparator treatments in three consecutive cycles. The trial drugs were etoricoxib 120 mg, lumiracoxib 400 mg, and rofecoxib 50 mg; comparators were placebo or naproxen. All studies were industry-sponsored.
• There were no significant differences in efficacy between the COX-2 inhibitors and standard NSAIDs in terms of pain relief. Both COX-2 inhibitors and standard NSAIDs were significantly more effective than placebo.
• The authors reported good tolerance to COX-2 selective inhibitors with few, but no serious, adverse events. The most frequent adverse events noted were dyspepsia, diarrhoea, gastritis, and vomiting in the RCT that compared mefenamic acid (common adverse effects [AEs] in 20%) with meloxicam 7.5 mg (common AEs in 6%) and meloxicam 15 mg (common AEs in 7%); and headache and nausea in the RCT that compared treatment with naproxen (common AEs in 25%), placebo (common AEs in 15%), and etoricoxib (common AEs in 12%).

Three additional RCTs (in two publications) have subsequently been published investigating valdecoxib and lumiracoxib for the treatment of primary dysmenorrhoea [Daniels et al, 2005; Daniels et al, 2008].

• These three double-blind crossover studies in a total of 397 women with primary dysmenorrhoea of moderate-to-severe intensity compared valdecoxib (20 mg or 40 mg twice a day) or lumiracoxib (200 mg once, or 200 mg when required) with placebo and naproxen.
• The results confirmed that all active treatments were superior to placebo, and that there was no evidence of a significant difference in pain relief provided by the COX-2 inhibitor (valdecoxib or lumiracoxib) and naproxen.
• All treatments were reported to be well tolerated.