DEAR DR. DONOHUE: I enclosed your recent column on PSA. I also have enclosed an article that sets forth the extremely low accuracy of the PSA and describes the highly accurate EPCA-2. I had a high PSA number, and my doctor was poised to poke and cut until I showed him the EPCA-2 article. He then informed me he could do the test in about two minutes, which he did. It came back negative. Why did you not offer more on the statistical failure of the PSA and not even mention the EPCA-2? — S.

ANSWER: No one, including me, represents the PSA (prostate specific antigen) as being a foolproof test for the detection of prostate cancer. The sensitivity of the PSA test — the reliability that the test will be positive if cancer is present — is in the range of 70 percent to 80 percent. The specificity of the test — the reliability that the test will be negative when there is no cancer — is 60 percent to 70 percent. It’s far from a perfect test. There is no perfect lab test for any illness. There is no perfect human being. The PSA is the best screening test we have. Gland enlargement, gland inflammation, infection and even bike riding can elevate the PSA.

Alternate tests can improve on the weaknesses of PSA. The increase in PSA value over time — PSA velocity — refines the test’s accuracy. Free-to-bound PSA is another way of improving the PSA’s predictability. Bound PSA is PSA attached to other blood substances. A low percentage of free PSA points to cancer.

The EPCA-2 test (early prostate cancer antigen-2) you referred to is not yet on the market. It holds the promise of greater accuracy. I checked with the office of the researcher who designed this test, and was told it may be another year before it appears for general use.

PSA testing leads to over- and under-diagnosing of prostate cancer. A positive test triggers the need for a prostate biopsy. A negative test lulls doctor and patient into thinking all is well when it might not be. All the same, the test has served a useful purpose.

The booklet on prostate cancer and enlargement discusses these topics in greater detail. To order a copy, write: Dr. Donohue — No. 1001, Box 536475, Orlando, FL 32853-6475. Enclose a check or money order (no cash) for $4.75 U.S./$6 Can. with the recipient’s printed name and address. Please allow four weeks for delivery.

DEAR DR. DONOHUE: My wrist on the thumb side hurts terribly and has hurt for 10 days. I didn’t injure it. Advil has helped the pain, but it’s still there. I am getting worried. Do you have any idea what this is? — G.P.

ANSWER: Let me have you perform a test. Fold your thumb onto your palm. Then fold your fingers over the thumb. Bend your wrist down and toward the little finger side of the hand. If this hurts, it’s a good bet you have de Quervain’s tendonitis. It’s an inflammation of tendons controlling thumb movement. It comes from overuse, usually from gripping actions where the thumb is tightly holding onto an object like a tennis racquet.

Ice the painful area for 15 minutes three times a day. If ice brings no relief, switch to heat. Get a thumb splint to rest the thumb. You can find one at a local drugstore. Continue with Advil if it’s helping.

If you don’t have a marked improvement in a week, you must see a doctor for proper diagnosis and treatment. A cortisone injection often brings quick relief.

DEAR DR. DONOHUE: For several years my wife had unpredictable, short episodes of a “pounding heart.” She had EKGs, but never during an episode. They were normal. Her heart doctor ordered a LifeWatch monitor that records heart action for three weeks. It showed that my wife had attacks of atrial fibrillation. She had to have a pacemaker and is now feeling better than she has in several years. — E.F.

ANSWER: Long-term cardiac monitoring has been a godsend for people who have transient attacks of heart symptoms, like your wife did. Your wife’s experience informs many people of a device that could be diagnostically lifesaving for them.

Dr. Donohue regrets that he is unable to answer individual letters, but he will incorporate them in his column whenever possible. Readers may write him or request an order form of available health newsletters at P.O. Box 536475, Orlando, FL 32853-6475. Readers may also order health newsletters from www.rbmamall.com.