The Quality and Outcomes Framework (QOF) is the world's largest healthcare experiment, handing out big cash incentives to doctors to prescribe more drugs to patients whom they diagnose with early signs of a chronic disease, from heart problems to diabetes.
Launched in the UK in 2004, it was the forerunner of other cash-for-performance schemes that have been set up around the world. There's a flawless logic to the schemes: drugs control diseases, and so diagnosing chronic diseases early enough—and prescribing drugs to treat them—puts a brake on their progress and takes pressure off intensive care services, such as hospitals, down the line.
The QOF started out by paying doctors around £15,000 ($19,400) a year as a cash incentive, but the sums have since escalated, and the scheme has so far cost £30 billion ($39 billion). Worth every penny, say the architects of QOF, who estimate it saves around 30,000 lives a year—and the same could be said for the other cash-for-performance schemes.
But does prescribing more drugs really save lives? Surprisingly, nobody had bothered checking, at least not until last year, when a group of researchers from the University of Michigan looked over the numbers. They explored the impact of the first seven years of the QOF on a group of 100,000 people in the UK, comparing their health and longevity to similar groups in countries that weren't operating cash-for-performance schemes and, as a result, weren't taking anywhere near as many pharmaceuticals.
To their surprise, the researchers discovered that the QOF wasn't making any difference whatsoever to mortality. In other words, people who weren't taking any drugs for chronic conditions such as ischemic heart disease (where arteries narrow and become blocked), hypertension (high blood pressure), stroke, diabetes, chronic kidney disease, asthma and obstructive pulmonary disease were living just as long.1
And there's a serious downside to taking so many drugs. The researchers didn't account for those who suffered a side-effect or adverse reaction to the many more drugs they were taking. Around 740,000 people die each year from a reaction to drugs, according to Xendo, a Netherlands-based pharmaceutical consultancy,2 while in the UK alone, 250,000 people need hospital care every year to treat an adverse reaction to a drug, some of whom presumably die.3
Paying doctors to prescribe more drugs is one of the drivers of a phenomenon called medicalization—or too much medicine, as the UK's British Medical Association (BMA) calls it. It's a catch-all term that includes overdiagnosis, unnecessary screening programs, a loosening of the definition of a disease—such as high blood pressure, where safe levels 20 years ago are now dangerous—and new technology.
This all adds up to overtreatment and escalating costs that benefit the doctor and the drug company, but rarely the patient. The BMA cites the example of thyroid cancer, which has seen a threefold increase in the number of cases detected in the last 30 years—mainly because of more sophisticated and sensitive testing—and yet the death rate has stubbornly remained at 0.5 deaths per 100 cases throughout that time.
Taking up the theme, a group of Australian researchers estimates that 500,000 people across 12 countries have been 'overdiagnosed' with thyroid cancer in the past 20 years, which has led to unnecessary surgery and lifelong medication.4
Overdiagnosis doesn't mean the patient never had a problem, but that it was relatively benign and wasn't life-threatening, explained Ray Moynihan from Bond University in Queensland, one of the researchers.
DCIS (ductal carcinoma in situ) is another example. It's an abnormality in the breast tissue that is often picked up by routine mammogram screening, and its discovery will trigger the full force of cancer therapy—chemotherapy, radiotherapy and even mastectomy, or breast removal.
But DCIS isn't cancer, and only in very rare cases does it develop into a tumor—yet women are forced through the pain and stress of cancer treatment. This is borne out by cancer statistics showing that, despite this overtreatment, the actual number of women who suffer advanced, life-threatening breast cancer hasn't changed in years.5
"There has traditionally been a sense that screening for cancers doesn't have a downside, but the evidence is suggesting screening healthy people can have downsides including overdiagnosis and overtreatment," said Professor Moynihan.6
His is far from an eccentric view. Australia's Cancer Council has issued a statement—which has been endorsed by Australasia's major medical groups—that calls for a rethink on medical care, and specifically the overdiagnosis and overtreatment that is harming patients.
In addition to DCIS, the council also includes prostate cancer as another example of overdiagnosis. Its chief executive, Professor Sanchia Aranda, said: "Men are diagnosed with a prostate cancer that may not even threaten their life, but in some instances can result in unnecessary or expensive treatment, causing a great deal of anxiety."7
Professor H. Gilbert Welch at the Dartmouth Institute for Health Policy and Clinical Practice agrees, and describes prostate cancer screening as the "poster child" for overdiagnosis. He says the PSA blood test is responsible "for over 1 million Americans being treated for a cancer that was never going to bother them."
In his book, Overdiagnosed: Making People Sick in the Pursuit of Health (Beacon Press, 2011), Professor Welch says that other conditions, such as hypertension, high cholesterol, diabetes and osteoporosis, have also fallen foul of overdiagnosis. "A fasting blood sugar of 130 was not considered to be diabetes before 1997, but now it is. And these numbers are always changing in one direction: the direction of labeling more and more people as abnormal," he said.8
Researchers at the University of Copenhagen—who describe overdiagnosis as "one of the most harmful and costly problems in modern healthcare"—have identified three drivers of overdiagnosis: overdetection, overdefinition and overselling.9
• Overdetection: The identification of abnormalities that were never going to cause harm.
• Overdefinition: Lowering the threshold that determines disease, such as the definition of hypertension, which changed from a blood pressure reading of more than 150 mmHg a few decades ago to today's 130 mmHg.
• Overselling: The creation of unpleasant everyday problems—such as sadness, sleeplessness or difficulty concentrating—into a chronic health problem, even though they are usually mild or fleeting.
And let's not forget the good old-fashioned profit motive, job justification, and medical and hospital targets, says Professor Welch. Drug companies, medical device manufacturers, imaging centers and even hospitals benefit from overdiagnosis, as does the doctor incentivized to prescribe more drugs. "The easiest way to make money isn't to build a better drug or device—it's to expand the market for existing drugs and devices by expanding the indication to include more patients."
There's also a human, and less sinister, aspect. Many doctors buy into the idea that more medicine is a good thing, and it's also the safer option. "While doctors can be punished for underdiagnosis, they are never punished for overdiagnosis," he said.
Yet, in the headlong rush for more medicine, just one person has been forgotten: the patient.
The unlucky 7
These are the diseases and conditions that are most commonly overdiagnosed: problems that usually won't harm someone, but nonetheless get the full medical treatment that may involve drugs or surgery.
1) Thyroid cancer. Sensitive screening tests—such as CT and MRI scans and ultrasound-guided aspiration—have tripled the number of cases of thyroid cancer being detected. Yet the death rate from the cancer has remained at 0.5 deaths per 100 cases—or five in every 1,000 people diagnosed—for the past 30 years. Modern screening technology has detected 500,000 new cases in the last 20 years, and these people are typically treated with chemotherapy and radiotherapy—but it hasn't made any difference to the numbers dying.1
2) ADHD. The number of children diagnosed with ADHD (attention deficit hyperactivity disorder) has risen by 50 percent in the US and has doubled in the Netherlands. Most of the children diagnosed are classified as having "mild to moderate" symptoms of ADHD—and since any diagnosis is subjective, there's every possibility they don't have the problem at all.
The definition of ADHD has widened in the last 10 years, and children who were once 'naughty' are now diagnosed. Overdiagnosis—and unnecessary medication—is costing the US around $500 million a year for a condition that is self-limiting: in other words, it often disappears as the child grows.2
3) Prostate cancer.It's a truth worth repeating: most men will die with prostate cancer, not from it. Estimates vary, but it's thought that between 20 and 50 percent of cases detected wouldn't have killed the patient. Nonetheless, men will have radical surgery or radiation, and around one-third of them suffer life-destroying side-effects, such as incontinence or loss of sexual function.
4) PCOS. Diagnosis of polycystic ovary syndrome is a classic example of overdefinition (see main article)—a lowering of the threshold where disease begins. In 1990, around 5 percent of young women were thought to have PCOS, but when the new definition was introduced in 2003, the rate rose to 21 percent.
Many women are being incorrectly labeled as having the problem, causing unnecessary worry about their fertility and long-term health, say researchers at the University of Sydney.3
5) Breast cancer. Between 19 percent and 30 percent of breast cancers picked up through routine mammogram screening are false positives—in other words, 'cancers' that aren't there. These include cases of DCIS (ductal carcinoma in situ), which, despite the name, isn't cancer but an abnormality.
Of the 60 percent of women whose breast cancer is detected by a mammogram, only between 3 and 13 percent are actually helped by the test.4
6) Mild hypertension.Up to 40 percent of adults have hypertension, or high blood pressure, and 22 percent have mild hypertension. Around half the people with mild hypertension—a systolic pressure of 140-150 mmHg—are now treated with drugs, even though there's no evidence to suggest it stops the problem from worsening or prevents death.
This overmedicalization of a benign problem has come about because the threshold has been changed. Once, hypertension started with a blood pressure reading of around 150, and today it is 130.5
7) Prediabetes. Almost overnight, around half the Chinese adult population was suddenly 'sick' with prediabetes, at least according to a change in its definition by the American Diabetes Association (ADA) in 2011.
Instead of impaired glucose tolerance, the ADA widened the definition of prediabetes to include elevated fasting glucose (blood sugar) levels. The result? A sudden epidemic of a non-disease requiring medication, say researchers from University College London.6