Fifteen years ago, I would have told you a blood test saved my father’s life. Now I’m not so sure.

In 1998, my father, Orest, broke the news to our family that he had prostate cancer. A new physician had ordered a prostate specific antigen, or PSA, test when my father turned 65, and subsequent tests showed the levels were slowly rising.

Ultimately he decided on surgery, and I recall how thankful I was for that little blood test that discovered the cancer early and in its most treatable phase. I had totally bought into the narrative. The surgery would give him years of additional life, allowing my father to attend my niece’s high school graduation, my brother’s wedding and scores of family holidays.

I’m still amazed at all he accomplished, earning a scholarship for college and medical school as a non-native English speaker with a hearing impediment. He spent 26 years as a doctor in the U.S. Army, including a tour of duty in Vietnam and stints in the Pentagon and Walter Reed Army Medical Center before retiring as a colonel.

Several years after his diagnosis, my father’s PSA levels rose again and his doctors recommended hormone therapy. By this time, he was in his late 70s and I had learned much more about the fallibility of the PSA test and the overtreatment of prostate cancer. I began to wonder whether this continued treatment was really worth it. Were we treating a life-threatening illness or something that grows so slowly that it might not affect him in any meaningful way?

My father was skeptical but relented to the treatments. He suffered through hot flashes and other side effects for several months before his numbers dropped again.

Two years later, his doctors recommended radiation treatment. This time I pushed harder. Was there any sense in further treatment? My father was nearly 80, on blood thinners after a valve replacement and having trouble with his kidneys. My mother asked his oncologist what would happen if my father declined.

“He will die of prostate cancer,” the doctor said flatly.

As a health journalist, I knew my father was unlikely to live much longer. Saying those words to your mother or siblings is much more difficult. It makes you sound uncaring to talk about statistics and life expectancy, instead of clinging to the myth that he’ll live forever.

Once again, my father relented — I suspect to appease the family. And again he suffered through side effects, but his PSA levels returned to near zero.

It’s so easy to conclude that a test that uncovers a cancer is a life-saver. We jump quickly to the assumption that our loved one would have died quickly if the cancer hadn’t been found and treated. It’s the narrative we’ve been fed for years and, perhaps, the narrative we’d most like to believe.

The truth is once a test comes back positive, it sets in motion a chain of events that shrouds our view of any other possible outcome. A positive test is virtually impossible to ignore. Not many can look cancer in the eye and not blink. And once someone undergoes treatment, it is impossible to know what would have happened otherwise.

Two years after his radiation treatments, my father fell down the stairs at his home in the suburbs of Philadelphia. He hit his head against the wall at the foot of the staircase, and with his blood thinned to prevent clotting, he bled to death at the age of 82.

I don’t know if my father would have died earlier without treatment, any more than his doctor “knew” he would die of prostate cancer without treatment. With repeated elevated PSA tests following treatment, his cancer may have been more persistent, more dangerous than average. I may have had the last 15 years with my father only because of his treatments. On the other hand, my father endured years of side effects. They were, in some way, the price he paid for our frailty, our inability to accept even the smallest odds that a father and mentor would be taken away from us any earlier than we had hoped. Whatever years we thought he was buying with his treatments vanished into mere conjecture one October afternoon.

I can’t help thinking whether any of it was worth it, whether I should have taken a stronger stand against further treatment. But it’s just as easy to say screening and treatment were futile after a patient dies, as it is to affirm the value of screening and treatment when the patient lives. It’s why screening and treatment recommendations must be based on overall statistics, not on individual stories or the ever-present, all-powerful anecdote.

In the end, there’s only one thing I know for certain: I miss my father terribly. •