Chandika 2012.

Methods	Study design: randomised controlled trial. Duration of study: October 2006 and March 2007.
Participants	Sample size: 65. Only 66% of the women presented with lactational breast abscesses diagnosed using ultrasound. Inclusion criteria: women 14 years and older with breast abscesses up to 5 cm in diameter as determined via ultrasound, who presented to the Accident and Emergency department and breast clinic with breast abscess. Exclusion criteria: women with recurrent or chronic breast abscesses and those with necrotic skin overlying the abscess or abscesses that were already draining. Women with clinical features of immune suppression (WHO clinical stage 111 and 1V) and those known to be allergic to penicillin and antibiotics were also excluded. Setting: Accident and emergency and breast clinic of Mulago Hospital complex, Kampala City. Country: Uganda.
Interventions	1. Ultrasound‐guided needle aspiration (n = 33) This was done in the department of radiology ultrasound room in the OPD. Aspiration was done using ultrasound guidance using local anaesthetic. Infected fluid samples were sent for culture and sensitivity. Aspiration was done until there was no infected fluid. Cloxacillin 500 mg orally 8 hourly for 10 days and Diclofenac 75 mg IM stat and 50 mg 8 hourly for 3 days respectively. Follow‐up was done via the OPD by the principal investigator on days 7, 14 and 30. If abscess persisted, aspiration was done on day 7, if it persisted on day 14 ‐ this was considered treatment failure. The women was then sent for I&D. Women were asked to resume breastfeeding on both breasts as soon pain during breastfeeding was tolerable. 2. Incision and drainage (n = 32) Procedure was done in the operating theatre under general anaesthesia. Women were hospitalised overnight and discharged the next day. The participant was placed in a supine position. The affected breast was swabbed using Chlorhexidine Centrimide. A skin incision was made at the area of maximum fluctuation along skin lines and a sinus forceps used to reach the abscess cavity. Infected fluid samples were sent for culture and sensitivity. The infected fluid was then evacuated and loculi broken down digitally, the wound was then packed with sterile gauze. Dicolfenac 75 mg injection (IM) stat and 50 mg 8 hourly for 3 days Cloxacillin 500 mg 8 hourly for 10 days. On discharge, wound dressings were done at nearby clinic until the wound healed. Women who were resistant to Cloxacillin were excluded from the study and antibiotics changed based on sensitivity studies. Women were asked to resume breastfeeding on both breasts as soon pain during breastfeeding was tolerable.
Outcomes	Authors did not differentiate between primary and secondary outcomes. Time to breast abscess resolution. Breast abscess recurrence. Acceptance of ultrasound‐guided needle aspiration procedure. Cost of the procedure. Definition of healing: healing was defined as achieving breast abscess resolution. Breast abscess resolution was defined as clinically no breast tenderness, swelling or wound at previous site of abscess and sonographically complete absence of fluid collection, normal breast glandular and fibro fat tissue with no edema.
Notes	Funding: not reported. Conflict of interest: the authors declared no competing interest. Ethics approval: approval issued by Faculty of medicine research committee, National science and research council, Mulago hospital complex and the Department of surgery, Mulago hospital. Author contacted: yes, no response received. Below are the questions sent to the authors via email: Is there a copy of the protocol available? If so we would be very grateful to get a copy. Can the authors also indicate what the numbers were lactating in the Ultrasound versus the I&D group included in the study. If numbers on parity (primi, multi or nulliparous) for the two groups are available are we able to request those as well? Is there any information on duration of breastfeeding in both groups, i.e. the number that were breastfeeding before the interventions and those that continued after and if so for how long? The review team also needs clarity about the following. The following are the numbers included in study; I&D N = 32 Ultrasound‐guided aspiration (UGA) = 33. Loss to follow‐up: I&D = 1; UGA = 4. The loss to follow‐up is taken from Figure 2. Therefore the numbers should be I&D group N = 31 and UGA group N = 29. In light of this can you please clarify the difference in the numbers in Table 1: the total healing rates per group: Group A = 29 and Group B = 28? The authors indicate that there was no conversion to I&D from the ultrasound group.  It was suggested that “Patients whose culture and sensitivity results showed resistance to Cloxacillin were excluded from the study and antibiotic treatment changed accordingly”. Can you advise how many patients were excluded from the study for ultrasound arm and did this apply to the incision and drainage arm as well?   Could the authors also clarify in Table 2 whether mean healing time was meant to read weeks or days? Lastly, the study indicated 100% maternal satisfaction with the ultrasound guidance –do you have the numbers for the Incision and drainage group?
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Microsoft Excel version 5 was used to generate a random number list.
Allocation concealment (selection bias)	Unclear risk	Not stated.
Blinding of participants and personnel (performance bias) All outcomes	Unclear risk	Although there was no blinding this would not affect the resolution of the abscess.
Blinding of outcome assessment (detection bias) All outcomes	Unclear risk	The authors reported that no blinding was done. However blinding of outcome assessors is not possible due to the nature of the interventions and therefore difficult to judge whether this has an impact.
Incomplete outcome data (attrition bias) All outcomes	Unclear risk	33 women were randomised to the ultrasound needle aspiration group with 32 allocated to the I&D group. The authors indicated that 29 were effectively treated by ultrasound‐guided needle aspiration group and 31 in the incision group with no reason being given. These numbers have not been included in the analysis and no ITT was done. Authors contacted and awaiting response.
Selective reporting (reporting bias)	Low risk	All outcomes prespecified in the methods section have been addresses.
Other bias	Unclear risk	Women in both arms were excluded from the study if they were shown to be resistant to cloxacillin after randomisation and it is not known how many women were resistant to cloxacillin in both arms. Authors declared no conflict of interest.