Gaitan 2002.

Methods	Design of RCT: parallel clinical trial                                                       Power calculation: stated                                                           No blinding used                                                           Follow‐up at one week                                                          110 participants recruited, 110 randomly assigned                                                              20 participants excluded before random assignment                                                            Lost to follow‐up: eight participants. No information about how they were distributed between groups                                          110 women analysed                                                    Single centre: Instituto Materno Infantil en Bogota, Colombia. A tertiary care maternity and gynaecological hospital                                                    Enrolment between November 1997 and June 2000                                                     Source of funding: Colombian Institute for the Development of Science and Technology, COLCIENCIAS   Ethical issues: protocol and written consent approved by the Universidad Nacional de Colombia Ethics Committee                                         Method used to establish definitive diagnosis: based on bacteriological cultures of endometrium and endocervix plus histopathological findings of endometrium plus visual examination in laparoscopic group and bacteriological cultures of endometrium and endocervix plus histopathological findings of endometrium plus visual examination or evolution of pain depending on whether participant had undergone laparotomy. Surgical samples sent for pathological study
Participants	Women between 18 and 45 years old who consulted because of non‐specific lower abdominal pain (NSLAP). NSLAP defined as one of the following: pain not proceeding in a classical course, or, after clinical history had been taken and physical examination, haemogram, urinalysis, pregnancy test and pelvic and transvaginal ultrasonography performed, two examiners did not agree on a diagnosis by the end of six hours of observation                                            Excluded: participants in whom pathology in upper hemi‐abdomen is suspected, with background of peritonitis or intestinal surgery, with two or more intra‐abdominal surgeries, with evidence of urinary infection, kidney lithiasis, cholelithiasis, infectious colitis or irritable colon, with multiple organic dysfunction syndrome, septic shock or hypovolaemic shock, with chronic pelvic pain or pain of more than three months’ evolution, with possible intrauterine pregnancy and unharmed sac, participants weighing more than 100 kg and those with psychiatric disorders                                          Mean age (SD): 27.6 years (± 6.7) in laparoscopy group, 30.2 years (± 6.7) in conventional group                                                     Mean temperature (SD): 36.7 °C (± 0.5) in laparoscopy group, 36.5 °C (± 0.9) in conventional group                                                       Mean serum white cell count (DS): 8771 (± 3418) in laparoscopy group, 10,253 (± 4029) in conventional group                                                            "Time of evolution of pain median (range): 3 days (1 ‐ 60) in laparoscopy group,  4 days (1 ‐ 80) in conventional group . of the pain (SD): 22.1 (± 5) in laparoscopy group, 21.5 (± 3) in the conventional group" Abdominal surgery background number of participants exposed (percentage): 27 (50.9%) in laparoscopy group and 20 (38.5%) in conventional group
Interventions	Laparoscopic diagnosis versus conventional diagnosis Laparoscopic diagnostic method defined as direct observation of the pelvic cavity with a Wolf laparoscope fitted with a video camera and zoom, a light source and the high‐flow insufflator that allows surgical procedures to be performed. IMI laparoscopy team is experienced in lower abdominal pain diagnosis. Laparoscopic diagnosis of inflammatory pelvic illness reached following Hager’s criteria. Diagnosis of ruptured ectopic pregnancy reached when a bluish mass was seen in the tube, whether associated with hemorrhagic material in the cul‐de‐sac or not Appendicitis and ovarian cyst diagnoses based on observing the changes described above visually. A healthy pelvis diagnosed when no alterations were found     Conventional diagnosis method based on clinical assessment and laboratory tests. It could have included surgical interventions, such as precision laparotomy, performed by the IMI emergency team. Diagnosis of inflammatory pelvic illness reached when at least two of Hager's main criteria were present. Diagnosis of appendicitis reached when signs of swelling or necrosis noted. Diagnosis of ectopic pregnancy reached by means of ultrasonography and serial human chorionic gonadotropin determinations, or laparotomy. Ovarian cyst suspected when a mobile mass was detected during pelvic examination, or when an ovarian mass larger than 5 cm was found with laparotomy. Cyst showing active haemorrhage or haemorrhagic content interpreted as ruptured. If adnexae were twisted, torsion diagnosis was made. Diagnosis of a healthy pelvis reached only when no alterations were found in the pelvic organs during laparotomy Both methods compared with a complex standard determined by both histopathological and microbiological criteria and evolution of the underling pathology. Surgical pathology also taken into account. Endocervical samples taken in 109 of 110 women assigned to one of the two groups. Gram stain and N. gonorrhoeae, C. trachomatis, and Mycoplasma cultures were done. Endometrial sample taken for culture of these and other aerobic and anaerobic bacteria with a Pipelle curette after washing of the exocervix with a saline solution. Another sample taken for histopathological examination. Positive endocervical culture or positive endometrial culture of STD bacteria interpreted as PID. PID diagnoses with endometrial biopsies reached using Kiviat’s criteria for endometritis. Decidua, an arias stella reaction or hypersecretory changes used to diagnose ectopic pregnancy or complicated ovarian cysts. Surgical pathology and endometrium biopsies read by one of the study authors
Outcomes	Conclusive diagnoses: number of cases in which a final diagnosis was reached in each method. Clinical observation, ultrasonography, para‐clinical exams, and visual findings at laparoscopy or laparotomy taken into account; stated in methods and reported                                                  Accurate diagnoses: accuracy calculated by comparing each method with a standard; stated in methods and reported                                                          Length of in‐patient stay before diagnosis: included only the time elapsed between hospital admission and the beginning of surgical or medical treatment, or spontaneous relief of pain. Not clearly validated; stated in methods and reported                                                              .            Procedural complications: caused by diagnostic intervention, delayed diagnosis or pain control before admission; stated in methods and reported
User defined 1
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Computer‐generated randomisation table
Allocation concealment (selection bias)	Low risk	Concealed in sealed, opaque, sequentially numbered envelopes
Blinding (performance bias and detection bias) All outcomes	High risk	Open RCT
Incomplete outcome data (attrition bias) All outcomes	Unclear risk	Not clear whether follow‐up was similar in the two groups. Eight of 110 lost to follow‐up
Selective reporting (reporting bias)	Unclear risk	Protocol of the study not available
Free of differential verification bias	Low risk	Definitive diagnosis done with the same reference standard in the two groups
Free of partial verification bias	Low risk	All participants undergoing the reference standard
Free of Incorporation Bias	High risk	Index test incorporated as the reference standard in the laparoscopic group
Other bias	Low risk	Baseline characteristics reported