Table 3 Feasibility results for this pilot review

Sensitivity/specificity^b

Six terms combined gave good sensitivity but compromised specificity

Published protocols^a

Few studies had pre-specified methods

Coefficient of variation^c

Fairly consistent number of databases searched

Justification of search limits^a

Few studies justified their search limits

Justification of sample size or perform sample size calculation^a

Few studies justified their sample sizes or performed calculations

Use a random sampling approach^c

All studies adopted a random sampling approach

Discuss the generalizability of findings^a

Few studies described how generalizable their findings were