Summary of findings'. 'Diagnostic laparoscopy.

Population	Males and females aged 15 to 87 years with potentially resectable pancreatic or periampullary carcinoma on computed tomography (CT) scanning
Setting	Surgical centres in the USA, Germany, the UK, Japan, Israel, and the Netherlands
Index test	Diagnostic laparoscopy with histologic confirmation
Reference standard	Paraffin section histology on diagnostic laparoscopy or laparotomy or surgeon's judgement of unresectability on laparotomy True positive: Suspicious lesion on diagnostic laparoscopy confirmed to be cancer by a histopathological examination of biopsy obtained during diagnostic laparoscopy False positive: This is not possible since laparotomy will only be performed if histopathology of the biopsy of the suspicious lesion on diagnostic laparoscopy shows no evidence of cancer False negative: No evidence of unresectability by diagnostic laparoscopy but evidence of unresectability on laparotomy True negative: No evidence of unresectability by diagnostic laparoscopy and laparotomy
Number of studies	16 studies
Summary sensitivity	64.4% (95% confidence interval 50.1% to 76.6%)
Consistent results	No
Uncertainty (overall risk of bias)	High
Other limitations	Different definitions of unresectability because studies used surgeon's judgement of unresectability on laparotomy when biopsy confirmation was not possible
Pre‐test probability from included studies1	Post‐test probability of unresectable disease for patients with a negative test result (95% confidence interval)2	Percentage of patients for whom unnecessary laparotomy can be avoided3
Minimum = 17.4	7.0 (4.9 to 9.8)	10.4
Lower quartile = 34.7	15.9 (11.4 to 21.6)	18.8
Median = 41.4	20.1 (14.7 to 26.8)	21.3
Upper quartile = 62.7	37.4 (29.0 to 46.6)	25.3
Maximum = 81.8	61.5 (52.3 to 70.0)	20.3
Interpretation	At pre‐test probabilities of 17%, 41%, and 82%, adding diagnostic laparoscopy to CT scan for the preoperative staging of pancreatic cancer avoids 10, 21, and 20 unnecessary laparotomies out of 100 laparotomies performed for curative resection purposes. These pre‐test probabilities are the minimum, middle, and maximum values obtained from the included studies

¹Probability of someone having unresectable disease at laparotomy after CT indicated that the disease is resectable.
 ²Probability of someone having unresectable disease after the CT and diagnostic laparoscopy indicated that the disease is resectable.
 ³Calculated as the difference between the post‐test probability and the pre‐test probability.

All probabilities are reported in the table as percentages.