Ketankumar Gajjar, Pierre PL Martin-Hirsch, Andrew Bryant, Gemma L Owens

Pre-cancerous lesions of cervix (cervical intraepithelial neoplasia (CIN)) are usually treated with excisional or ablative procedures. In the UK, the National Health Service (NHS) cervical screening guidelines suggest that over 80% of treatments should be performed in an outpatient setting (colposcopy clinics). Furthermore, these guidelines suggest that analgesia should always be given prior to laser or excisional treatments. Currently various pain relief strategies are employed that may reduce pain during these procedures. To assess whether the administration of pain relief (analgesia) reduces pain during colposcopy treatment and in the postoperative period. We searched the Cochrane Central Register of Controlled Trials (CENTRAL 2016, Issue 2), MEDLINE (1950 to March week 3, 2016) and Embase (1980 to week 12, 2016) for studies of any design relating to analgesia for colposcopic management. We also searched registers of clinical trials, abstracts of scientific meetings, reference lists of included studies and contacted experts in the field. Randomised controlled trials (RCTs) that compared all types of pain relief before, during or after outpatient treatment to the cervix, in women with CIN undergoing loop excision, laser ablation, laser excision or cryosurgery in an outpatient colposcopy clinic setting. We independently assessed study eligibility, extracted data and assessed risk of bias. We entered data into Review Manager 5 and double checked it for accuracy. Where possible, we expressed results as mean pain score and standard error of the mean with 95% confidence intervals (CI) and synthesised data in a meta-analysis. We included 19 RCTs (1720 women) of varying methodological quality in the review. These trials compared a variety of interventions aimed at reducing pain in women who underwent treatment for CIN, including cervical injection with lignocaine alone, lignocaine with adrenaline, buffered lignocaine with adrenaline, prilocaine with felypressin, oral analgesics (non-steroidal anti-inflammatory drugs (NSAIDs)), inhalation analgesia (gas mixture of isoflurane and desflurane), lignocaine spray, cocaine spray, local application of benzocaine gel, lignocaine-prilocaine cream (EMLA cream) and transcutaneous electrical nerve stimulation (TENS).Most comparisons were restricted to single trial analyses and were under-powered to detect differences in pain scores between treatments that may or may not have been present. There was no difference in pain relief between women who received local anaesthetic infiltration (lignocaine 2%; administered as a paracervical or direct cervical injection) and a saline placebo (mean difference (MD) -13.74; 95% CI -34.32 to 6.83; 2 trials; 130 women; low quality evidence). However, when local anaesthetic was combined with a vasoconstrictor agent (one trial used lignocaine plus adrenaline while the second trial used prilocaine plus felypressin), there was less pain (on visual analogue scale (VAS)) compared with no treatment (MD -23.73; 95% CI -37.53 to -9.93; 2 trials; 95 women; low quality evidence). Comparing two preparations of local anaesthetic combined with vasoconstrictor, prilocaine plus felypressin did not differ from lignocaine plus adrenaline for its effect on pain control (MD -0.05; 95% CI -0.26 to 0.16; 1 trial; 200 women). Although the mean (± standard deviation (SD)) observed blood loss score was less with lignocaine plus adrenaline (1.33 ± 1.05) compared with prilocaine plus felypressin (1.74 ± 0.98), the difference was not clinically as the overall scores in both groups were low (MD 0.41; 95% CI 0.13 to 0.69; 1 trial; 200 women). Inhalation of gas mixture (isoflurane and desflurane) in addition to standard cervical injection with prilocaine plus felypressin resulted in less pain during the LLETZ (loop excision of the transformation zone) procedure (MD -7.20; 95% CI -12.45 to -1.95; 1 trial; 389 women). Lignocaine plus ornipressin resulted in less measured blood loss (MD -8.75 ml; 95% CI -10.43 to -7.07; 1 trial; 100 women) and a shorter duration of treatment (MD -7.72 minutes; 95% CI -8.49 to -6.95; 1 trial; 100 women) than cervical infiltration with lignocaine alone. Buffered solution (sodium bicarbonate buffer mixed with lignocaine plus adrenaline) was not superior to non-buffered solution of lignocaine plus adrenaline in relieving pain during the procedure (MD -8.00; 95% CI -17.57 to 1.57; 1 trial; 52 women).One meta-analysis found no difference in pain using VAS between women who received oral analgesic and women who received placebo (MD -3.51; 95% CI -10.03 to 3.01; 2 trials; 129 women; low quality evidence).Cocaine spray was associated with less pain (MD -28.00; 95% CI -37.86 to -18.14; 1 trial; 50 women) and blood loss (MD 0.04; 95% CI 0 to 0.70; 1 trial; 50 women) than placebo.None of the trials reported serious adverse events and majority of trials were at moderate or high risk of bias (13 trials). Based on two small trials, there was no difference in pain relief in women receiving oral analgesics compared with placebo or no treatment (MD -3.51; 95% CI -10.03 to 3.01; 129 women). We consider this evidence to be of a low to moderate quality. In routine clinical practice, intracervical injection of local anaesthetic with a vasoconstrictor (lignocaine plus adrenaline or prilocaine plus felypressin) appears to be the optimum analgesia for treatment. However, further high quality, adequately powered trials should be undertaken in order to provide the data necessary to estimate the efficacy of oral analgesics, the optimal route of administration and dose of local anaesthetics.