Thursday, November 19, 2009

The Evidence for Evidence-based Medicine

The recent recommendations of the US Preventive Services Task Force (USPSTF) regarding mammography have created confusion and resistance from both the medical community and the general public. The USPSTF recommendations were based on the application of evidence-based medicine (EBM). The current firestorm threatens the entire EBM enterprise.

My recommendation to Kathleen Sebelius, the US Secretary of Health and Human Services (who walked away from the task force recommendations), is to take a cue from the Intergovernmental Panel on Climate Change (IPCC). Ask the USPSTF to stick to the science and avoid the policy recommendations.

The EBM enterprise essentially conducts a literature review of existing studies in the field and assesses the risks and benefits of a particular treatment. This is essentially what the IPCC does in it's Assessment Reports (ARs) and this is where the IPCC scientists stop, understanding that they are not politicians.

EBM as practiced by USPSTF (I know this is acronym overload, but can't help myself), however, can't resist taking two further steps. Evidence-based guidelines (EBG), meant to be applied at the organizational or institutional level, use EBM literature reviews to develop guidelines, policy and regulations for medical practice. And, USPSTF appears willing to even take a further step into evidence-based individual decision making (EBID) which is the application of evidence-based medicine to the individual patient.

Purely from the standpoint of statistics, EBID is a dangerous idea. Population statistical data and results have little if any useful application to individuals. All population statistics describe central tendencies and confidence intervals based on distributions of patients. There is simply no way for a clinical practitioner to locate an individual patient within that data storm.

Even though EBID is not useful, the literature reviews provided by EBM are very useful. The medical literature is huge and of widely varying quality. USPSTF provides an extremely useful service reviewing this research and reaching a consensus opinion on the risks and benefits of a particular treatment. As should be obvious from the mammography debate, they loose their credibility by going any further. Let's shift the pressure to the organizations and institutions that deliver health care to use the gold-standard EBM results to develop guidelines, policies and regulations. And, let's require them to defend those decisions with more outcomes data.

1 comment:

  1. The recommendations of the USPSTF included exceptions to the general guidelines for women with a family hx of breast cancer or other risk factors specific to the individual.

    All of the objections to the USPSTF, or at least those I saw, were lodged by the Radiologists, who shouldn't be ignored, but do stand to lose income if they were adopted. They didn't think to mention that in their remarks. Gordon Ridley

    ReplyDelete