Miguel Maldonado Fernandez MD PhD MSc
Laura Rubio Rodriguez MD MSc
The name “Hawthorne effect” derives from Western Electric Hawthorne Works, a plant near Chicago in which Elton Mayo, Professor of Industrial Management at the Harvard Business School, and his assistant Fritz Roethlisberger carried out behavioural experiments to measure the outcome of the workers. Roethlisberger defined the Hawthorne effect is the fact that if you know that you are enrolled in a behavioural study, you will likely change your performance in response to being observed1.
From our personal experience and communication with other colleagues, it is usually argued that blinding is a solution for the Hawthorne effect in a clinical trial. However, we do not thing that is the case. In a trial, both allocation groups need to be consulted to request the informed consent2,3. Therefore, for ethical reasons it is mandatory that anyone enrolling in a clinical trial must be aware of the fact that they are being “observed”. On the other hand, blinding implies that participants (in the case of single blinding) does not know if they are receiving the active intervention or the control. But they know that there is a trial going on.
Some others have argued (also from personal communication) that randomization controls the Hawthorne effect. Indeed, if we reckon the fact that both groups of the trial are subject to the Hawthorne effect, we can conclude that randomizing would have the same result as for other confounders, like age, or smoking habits (i.e. if both groups have the same age distribution or smoking habits, these variables will not act as confounders). Therefore, although we can admit that randomization controls the Hawthorn effect between groups, we must still accept that it is still active and working in all the participants.
Rapid Response:
Re: Understanding the Hawthorne effect
Hawthorne effect, blinding and randomization
Miguel Maldonado Fernandez MD PhD MSc
Laura Rubio Rodriguez MD MSc
The name “Hawthorne effect” derives from Western Electric Hawthorne Works, a plant near Chicago in which Elton Mayo, Professor of Industrial Management at the Harvard Business School, and his assistant Fritz Roethlisberger carried out behavioural experiments to measure the outcome of the workers. Roethlisberger defined the Hawthorne effect is the fact that if you know that you are enrolled in a behavioural study, you will likely change your performance in response to being observed1.
From our personal experience and communication with other colleagues, it is usually argued that blinding is a solution for the Hawthorne effect in a clinical trial. However, we do not thing that is the case. In a trial, both allocation groups need to be consulted to request the informed consent2,3. Therefore, for ethical reasons it is mandatory that anyone enrolling in a clinical trial must be aware of the fact that they are being “observed”. On the other hand, blinding implies that participants (in the case of single blinding) does not know if they are receiving the active intervention or the control. But they know that there is a trial going on.
Some others have argued (also from personal communication) that randomization controls the Hawthorne effect. Indeed, if we reckon the fact that both groups of the trial are subject to the Hawthorne effect, we can conclude that randomizing would have the same result as for other confounders, like age, or smoking habits (i.e. if both groups have the same age distribution or smoking habits, these variables will not act as confounders). Therefore, although we can admit that randomization controls the Hawthorn effect between groups, we must still accept that it is still active and working in all the participants.
References:
1. The Hawthorne Effect. Baker Library. Historical Collections. Harvard Business School. Available online at: [http://www.library.hbs.edu/hc/hawthorne/09.html].
2. WMA Declaration of Helsinki - Ethical Principles for Medical Research Involving Human Subjects. 64th WMA General Assembly, Fortaleza, Brazil, October 2013. Available online at [http://www.wma.net/en/30publications/10policies/b3/].
3. International Conferences on Harmonization. ICH E6. Available online at [http://www.ich.org/products/guidelines/efficacy/efficacy-single/article/....
Competing interests: No competing interests