TY  - RPRT
TI  - Reasons for caution when evaluating health care interventions using non-randomised study designs.
AU  - B C Reeves
PY  - 2004
DO  - 10.1159/000080575
UR  - https://pubmed.ncbi.nlm.nih.gov/15353902/
ID  - 15353902
ER  -