Jhee 2004.

Methods	Double blind randomised controlled trial, 4 weeks
Participants	21 participants (16 intervention, 5 control) who had a diagnosis of probable AD. Subjects had to meet the NINCDS‐ADRDA and DSM4 criteria for probable AD, had modified Hachinski Ischemia Scale score of less than or equal to 4, MMSE score between 10‐24 inclusive and had been diagnosed with probable AD for at least a six month period. They were out‐patient, aged>/=60 y in USA
Interventions	Interventions: Celecoxib dose 50. 200 and 400 mg oral twice Control: Placebo
Outcomes	Cognition (ADAScog 0‐70 lower score = improvement) at 4 weeks (MMSE higher score = improvement) at 4 weeks
Notes	Sources of funding not stated
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	High risk	Quote: Subjects with AD were randomly allocated a treatment arm in the order in which they were enrolled in the study Comment: Sequence generated by some rule based on the number they enrolled in the study
Allocation concealment (selection bias)	High risk	No information Comment: Probably not done
Blinding (performance bias and detection bias) All outcomes	Low risk	Quote 1: All study medication had the same appearance and was provided as celecoxib 50 mg and 200 mg capsules and matching placebo. Comment: Probably done
Incomplete outcome data (attrition bias) All outcomes	Low risk	Comment: Reasons for missing outcome data unlikely to be related to true outcome
Selective reporting (reporting bias)	Low risk	Comment: The study protocol is not available, but all of the study’s pre‐specified outcomes have been reported in the pre‐specified way in methodology
Other bias	High risk	Comment: The study appears to have some degrees of selective bias as patients' age in each group were difference.