Using Best-Worst Scaling to Understand Patient Priorities: A Case Example of Papanicolaou Tests for Homeless Women

Dear authors

I came across your article this week. I wanted to learn more about the BWS approach in eliciting patient preferences and a colleague recommended this article to me.

I was dumbstruck after reading it.

Firstly, I thought it odd that the authors believe that the higher risk of mortality is of major concern to homeless women. Homelessness is often the end of a long journey full of traumatic experiences. Nobody choses to be homeless. The implications of no longer having an address are too drastic. Homeless women were often exposed to violence and abuse. Yes, homeless women are vulnerable. So is everybody without shelter and protection. However, they are vulnerable because we as a society do not provide income support, accommodation or universal access to healthcare to such vulnerable members of our society, to members in need. The stated elevated risk of cervical cancer as a concern of homeless women sounded more like an excuse to conduct research in these women, not because it is a top concern to them.

Secondly, I would object to characterizing homeless women as “vulnerable population”. Women who are homeless, to my knowledge, do not share a unique set of individual characteristics that can be used to identify a sub-group of the overall population. Rather, they experienced one or several, sometimes traumatic, events that led, ultimately, to homelessness. So they share an experience, not a set of characteristics.

Lastly, the fact that the authors decided to conduct research in women who clearly are in a vulnerable situation to test a survey method shook and shocked me. The shock turned to anger upon realizing that the results of the research is unlikely to benefit homeless women. In fact, the overall benefit of this study is unclear. To assess patient preferences in patients for a PAP test among women who, I assume, prefer not to be homeless is like trying to find an answer for the wrong question. I am a man and I have never been homeless. So I do not know the preferences of homeless women. But the authors did not make a clear case why, while being homeless, getting a PAP test is of relevance to homeless women. The situation of homeless women is unlikely to change as a result of the study or the compensation of 10 dollar. I am not suggesting homeless women should not be the focus of research. I am advocating that research in currently vulnerable members of our society should benefit them.

I hope that in the future we use the intelligence, energy and financial resources of an institution like Harvard to advocate for a society where vulnerable members are sheltered, taken care of and helped so they are no longer homeless.