Patients become frustrated when their doctors do not see things as they do. So what happens when doctors taking care of them and researchers studying them do not see things the same as each other?

Consider chronic prostatitis, an illness whose treatments are often ineffective. To develop new knowledge that may permit better treatments, researchers carry out studies of patients with chronic prostatitis. As a necessary step, researchers must standardize symptom measures: They must compare "apples to apples." Prostatitis researchers most commonly use the NIH-Chronic Prostatitis Symptom Index (CPSI) to specifically measure the following dimensions of illness at a specific point in time:

• Pain or discomfort relating to the perineum, penis, testes, and lower abdomen
• Urinary symptoms
• The impact of the symptoms on quality of life

How well do CPSI's measurements match doctors' measurements? To answer this question, a group of scientists in Minnesota conducted a survey of men aged 47 to 90 years (median age 61.5).  Among their findings was a poor correlation between CPSI's specific measurements — pain in the penis and testes; pain on ejaculation — and a doctor-made clinical diagnosis of prostatitis.

Why was there a poor correlation?

There are many possible explanations for the study observations. Perhaps the most simple is that in sending out a questionnaire apart from the doctor visit the study was measuring symptoms that differed from those at the time of the visit. In other words, given that prostatitis symptoms vary with time, the pinpoint measurements may simply have missed them.

Another potentially imporant confounder is that the study population was generally older than other prostatitis study populations. In the older men, there is likely to be a high prevalence of benign prostate hyperplasia (BPH), with symptoms partly overlapping those of men with prostatitis. In other words, the study population, arguably with relatively high BPH prevalence may have become somewhat confused about the symptoms being queried.

Another limitation is that there really is no single absolute definition of prostatitis, such that the CPSI and doctors are actually measuring different things; this is especially is, as is likely to be the case, doctors are not using absolutely uniform diagnostic criteria. Notice, for example, how one online prostatitis consultation form asks about variables that are specifically not included in the CPSI; other providers may ask somewhat differently.

Among the things doctors might use to diagnose prostatitis that the CPSI does not use are:

• Queries relating to symptom severity over time
• Queries about such related symptoms as back pain
• Queries about such other dimensions as fatigue
• Physical examination
• Laboratory tests, such as a urinalysis

So is the CPSI a prostate diagnostic tool? This study demonstrates that it is not and should probably not be used as a sole element in bedside clinical assessment.

Is the CPSI useless? Not at all. The CPSI is an important tool in prostate research exactly because it standardizes measurement from patient to patient, from point A to point B, from treatment A to treatment B, from group to group. In research, it is required, but apparently insufficient.

If you think you have prostatitis, don't rely on a copy of the CPSI for diagnosis. Instead, ask your doctor for help.