You feel better but your doctor tells you that according to a laboratory test, you are still sick. Who do you believe? A recent randomized trial compared outcomes in four groups of patients suffering from asthma:

1. albuterol inhaler,
2. placebo inhaler,
3. sham acupuncture (a needle that either avoids known ‘acupuncture’ points or does not penetrate the skin), and
4. no intervention (a waiting list).

Subjective patient reports indicated that the albuterol, placebo, and sham acupuncture all reduced subjective reports of wheezing and feeling suffocated equally. That means if you have asthma, you feel as much better after taking a ‘placebo’ or the ‘real’ drug. However, on the ‘objective’ outcome, how much air you can blow in 1 second (the ‘FEV’) the real albuterol inhaler outperformed the placebo inhaler and sham acupuncture.

Placebo sceptics might argue that patients in the placebo and sham acupuncture groups merely reported that they felt better, to please the investigator (this is sometimes called ‘reporting bias’). However, this is unlikely because the patients receiving no intervention also reported improvements and presumably they had no reason to please the investigators (who after all had ignored them!)

Many prefer ‘objective’ outcomes because they believe that they address the underlying cause. But as Dan Moerman points out, patients suffering from asthma visit the doctor because they wheeze and feel suffocated not because they suffer from a reduced FEV. Indeed for many conditions including migraine, schizophrenia, back pain, depression, asthma, post-traumatic stress disorder, neurologic disorders such as Parkinson's disease, inflammatory bowel disease and indeed any condition defined by symptoms, requires that patient-preferred (usually subjective) outcomes trump objective outcome measures.

More generally, what patients, doctors, and policymakers care about (or should care about) is whether a medical intervention makes a patient live better or live longer. Many ‘objective’ outcome measures such as FEVs, cholesterol levels or arrhythmias are, in fact, surrogates for patient-relevant outcomes. While these surrogates are often extremely helpful and can lead to living better or longer, the links to the patient-relevant outcome is often far more spurious than is generally assumed.

So, which outcomes should you trust? It depends on your point of view …