Types of studies

We will include only randomized, controlled trials (RCTs) or quasi‐randomized trials. The RCT is the most appropriate method of study design, especially in the setting of analgesia, as suggestibility, patient expectations and the placebo effect are potentially significant sources of bias (Jadad 2000).

We will include trials that were recently published in abstract format or presented at national or international society meetings.

We will attempt to locate unpublished studies by contacting relevant manufacturers and investigators. We will not consider data from review articles, case reports or letters to the editor.

Types of participants

We will include adult and paediatric patients of either gender. In the paediatric group we will consider patients younger than 12 years of age. Recognizing that the lower age limit at which children can credibly quantify pain intensity is controversial (Tyler 1993), we will not set a minimum age threshold, so that we can identify as many relevant studies as possible.

Types of interventions

We will include only trials that evaluate the efficacy of topical local anaesthetics for repair of dermal (skin) lacerations. We will include only comparisons between:

1. infiltrated local anaesthetic agents with topically applied local anaesthetic agents;

2. different topical local anaesthetic formulations.

We will include both amide and ester local anaesthetics. We will accept topical preparations that contain more than one different local anaesthetics. We will also include multi‐component topical anaesthetics that contain vasoconstrictors (i.e., cocaine, epinephrine). We will define topical anaesthetics as any agent that is directly applied to the skin. Acceptable forms of topical local anaesthetics include liquid solution, gel, cream, ointment, lotion, jelly, balm, or aerosol spray. We will exclude studies that administer local anaesthetics using iontophoresis (a mild electrical current).

We will exclude papers investigating topical anaesthesia of mucous membranes (moist lining of the mouth, nose, eyes). In order that the procedures to be evaluated involve an approximately equivalent intensity and quality of pain, we will limit the technique of skin closure to instrumentation that involves piercing the skin with a needle (i.e., suture placement or stapling). Less invasive approaches to repair lacerations, such as application of tape or tissue adhesives, will be excluded. We will include only superficial injuries involving the epidermis or dermal layers. We will not consider deeper wounds involving the fascia or non‐skin structures. We will set no limitations on the length of the laceration, but we will exclude procedures on infected wounds. We will also exclude studies that administered systemic analgesics or sedatives that can alter the accuracy of the patients' perceived pain intensity.

Types of outcome measures

The primary outcome measured will be analgesic efficacy, reflected in the patient's self‐report of pain intensity during repair of the wound. Acceptable tools to quantify pain intensity include the visual analogue scale (VAS), numeric rating scale, verbal rating scale, faces scale, or other validated descriptors of pain intensity. Non‐concordance has been demonstrated between patients' and practitioners' assessments of procedure‐related pain (Singer 1999). Therefore, we will consider surrogate pain scores provided by the physician, parent or other observers only in instances where patient‐rated pain intensity is not reported.

We will extract other secondary outcomes, including the incidence of topical anaesthetic failure, requirement for supplemental anaesthesia, patients' acceptance of anaesthesia or patient behavioural responses, from studies when reported.

We will also note topical anaesthetic‐related acute toxicity or adverse effects. We will extract untoward effects including allergic reactions, neurological and cardiovascular toxicity. We will not include data concerning chronic complications, including wound infections or the adequacy of wound healing.

Electronic searches

We will search the Cochrane Central Register of Controlled Trials (CENTRAL, The Cochrane Library); MEDLINE (1966 to date); EMBASE Drugs and Pharmacology (1980 to date); CINAHL (1982 to date). For detailed search strategy see Appendix 1.

Searching other resources

We will manually search the following journals: Academic Emergency Medicine, Annals of Emergency Medicine, Emergency Medicine Clinics of North America and Journal of Emergency Medicine, Emergency Medicine Australasia (formerly known as Emergency Medicine) (1980 to date). We will search for additional papers by manually searching the references of any relevant articles found in these journals.

We will manually review recent abstracts presented at relevant national or international society meetings (February 2002 to date). We will contact the primary investigators to obtain any incomplete data. Furthermore, we will attempt to locate unpublished studies by directly contacting pertinent manufacturers and investigators.

We will implement no language restrictions in the literature search.

Assessment of studies for inclusion in the review

Two authors (AE, IE) will independently review the study titles and the abstracts identified by the search strategy. We will obtain the full reference if at least one author decides that the study may potentially meet the inclusion criteria. Two authors (AE, IE) will independently examine the complete articles to select the trials that meet the inclusion criteria. In the event of disagreement, a third author (DC) will be consulted."

Data extraction

Two authors (AE, IE) will independently read the articles and extract the data; we will compare the retrieved information and verify for accuracy. We will resolve disagreements by consensus. For studies that present the results only in graph format, two authors (AE, IE) will independently extract numerical data according to measurement of the graphs with a ruler. The average of the two measurements will then be calculated.

We will extract the following data:

1. study size (number of patients);

2. setting of study (single versus multi‐centre, inpatient versus outpatient);

3. patient characteristics (age range, gender distribution);

4. wound characteristics (length, location of laceration);

5. infiltrated and topical anaesthetics used in the study;

6. primary measures of pain intensity (patient's self‐report of pain intensity using VAS, numerical rating scale, or other valid measures of pain intensity);

7. secondary measures of pain intensity (incidence of topical anaesthetic failure, requirement for supplemental anaesthesia, patient behavioural responses, patients' acceptance of anaesthesia, surrogate pain scores etc.);

8. topical anaesthetic related acute toxicity or acute adverse effects;

9. administration of systemic analgesics or sedatives;

10. method of study randomization and double blinding;

11. study withdrawals and dropouts;

12. nature of allocation concealment.

Assessment of methodological quality

Two authors (AE, IE) will independently evaluate the methodological quality of each study. We will assess the most important aspects of high‐quality RCTs, including appropriate randomization, adequate double‐blinding, allocation concealment and description of withdrawals or dropouts. However, we will not use a specific grading system to calculate a methodological quality score.

Data analysis

If the reported outcomes (pain intensity measures) in the studies of the same anaesthetic agent are statistically combinable, then we will perform a meta‐analysis. We will compute calculations with Review Manager 4.2 (Rev Man 4.2). We will use the mean pain scores and standard deviations to derive a weighted mean difference (WMD), as well as the 95% confidence intervals. We will compute a chi‐squared test to test for heterogeneity. If there is heterogeneity, then we will use a random effects meta‐analysis. However, if the diversity between trials is significant, we will consider a qualitative systematic review. Furthermore, if the outcomes reported are not statistically combinable we will then perform a qualitative analysis.

We will analyse dichotomous data using Review Manager 4.2 (Rev Man 4.2). We will calculate the relative risk and the risk difference. If possible, we will perform a subgroup analysis to determine whether there are different results between adult and paediatric patients. We will consider patients younger than 12 years old in the paediatric group and allocate patients 12 years or older to the adult subgroup.

No sensitivity analysis is planned for this review.

We will first present the evidence regarding cocaine‐containing topical local anaesthetics. This will include comparisons between topical anaesthetics and infiltrated local anaesthetics, as well as comparisons between different formulations of cocaine‐based topical anaesthetics. Next, we will summarize the evidence concerning cocaine‐free topically applied local anaesthetics. Similarly, we will compare cocaine‐free topical agents and infiltrated local anaesthetics. Moreover, we will compare cocaine‐free topical anaesthetics with both cocaine‐based topical anaesthetics and different formulations of cocaine‐free topical anaesthetics.