Levels of Evidence
Systematic Review of Cohort Studies

Example: Looking at the charts of 100 patients in a clinic who smoke and comparing the rates of cancer with 100 other patients from the same clinic who don’t smoke.

Advantages:

• Can be performed in areas such as the effects of smoking, where ethical concerns would prevent the performance of a prospective RCT
• Less expensive than performing RCTs
• May improve the validity of conclusions from individual cohort studies

Disadvantages:

• It is extremely difficult to control for all potential biases. In theory, randomization controls for biases that the investigators are not even aware of.
• Just like a systematic review, if the studies that are combined for analysis aren’t homogeneous, the validity of the conclusions may be called into question.
• As with individual cohort studies, it is very difficult to control for all biasing influences (especially the ones unknown to the investigators). An excellent example of this is the results of the Women's’ Health Initiative where all previous studies were cohort studies that showed a benefit in reduction of heart disease from using HRT. The WHI was the first prospective RCT and had completely opposite results than the previous cohort studies. The “well woman bias” or the fact that most of the women in the cohort studies who received HRT elected to take HRT was not controlled for. In the cohort studies, women who elected to take HRT were just generally healthier than the general population because they were generally more proactive at preventative health measures. When the RCT was performed, this bias was controlled for by randomization because the decision of whether or not to take the HRT was randomized.