The well are being harmed by medicine that is overdiag-nosing, detecting illnesses that aren’t there and changing the boundaries of what illness is. Up to a third of people—who, a generation before, would have been considered healthy—are today taking drugs or undergoing dangerous treatments they don’t need.
When once medicine prided itself on treating the sick, it is today almost as likely to harm the healthy through drug overdose, overtreatment and overdiagnosis. Mass screening programmes that use increasingly sensitive technol-ogy are detecting cancers that will never endanger the patient and ‘abnormalities’ that are benign, while definitions of ‘disease’—such as blood pressure and cholesterol levels—are becoming tighter, thus bringing millions of people into the ‘disease net’. Abnormalities that will never develop into something that will affect the patient, or endanger life, are called ‘pseudodiseases’ (Arch Intern Med, 2011; 171: 1268–9).
Definitions of disease are driven by a desire to catch problems early—and for finding more customers for drugs. Many of the regulators who sit on boards that determine when disease begins have direct links to pharmaceutical companies that stand to benefit from a larger market for their drugs, says medical researcher Ray Moynihan, from Bond University
in Queensland, Australia (BMJ, 2012; 344: doi: 10.1136/bmj.e3502).
One example of the changing fashions for determining disease is the T score—which measures bone density—in menopausal women to determine the risk of osteoporosis (brittle bones). The average T score is set at zero and is based on a healthy, premenopausal woman, so any older woman is likely to have a negative T score. The World Health Organization (WHO) arbitrarily set the T score for osteoporosis at -2.5, but this was changed by the US National Osteoporosis Foundation to -2.0 in 2003—and, quite literally overnight, another 6.7 million American women had osteoporosis when just the day before they thought they were healthy.
The same happened a decade earlier when medicine tightened its definitions for cardiovascular diseases and, in particular, blood pressure and cholesterol levels. In 1997, the ‘danger’ signal for high blood pressure (hypertension) was changed from a reading of 100/ 160 mmHg to 90/140 mmHg, thereby suddenly making 13 million Americans candidates for antihypertensive drugs.
A further 42 million Americans overnight found themselves in line for a statin drug when the ‘harmful’ level of cholesterol in the blood was reduced from 240 mg/dL to greater than 200 mg/dL. As most healthy people have a cholesterol level of around 200 mg/dL, this casts the disease net very wide.
Aggressive overmedicalization comes at a cost both in human and financial terms. Fiscally, it’s reckoned that the United States alone is spending around $200 billion (£128 billion) every year on unnecessary treatments. Yet, the human cost is far higher. Many hundreds of thousands of people experience adverse reactions to antihypertensives, statins and other drugs, while overdiagnosis is also taking a heavy toll, from full mastectomies (complete removal of the breast) for women who don’t even have cancer to life-destroying treatments for men who would have died with prostate cancer, and not from it.Overdiagnosis overload
At least a third of people taking a drug or receiving treatment are healthy. This is an average figure and can be alarmingly far higher for some ‘diseases’, such as high cholesterol, prostate cancer, ADHD (attention-deficit/hyperactivity disorder) in children and adolescents, and asthma.
- Breast cancer. Up to 54 per cent of women who are diagnosed with breast cancer after a mammogram don’t have the disease but, instead, probably have DCIS (ductal carcinoma in situ), which develops into cancer in only around 10 per cent of cases. Nonetheless, surprisingly few oncologists recognize that DCIS is invariably benign and, instead, prescribe a range of aggressive responses up to full breast removal (Lancet Oncol, 2007; 8: 1129–38). Other studies have produced lower overdiagnosis rates, but they are still high, ranging from 52 per cent (BMJ, 2009; 339: b2587) down to 30 per cent (Cancer Causes Control, 2010; 21: 275–82) and 25 per cent (Ann Intern Med, 2012; 156: 491–9).
- Thyroid cancer. The chances of getting a thyroid cancer diagnosis wrong are very high. Sensitive screening technology often detects an abnormality in the thyroid that will never develop into cancer. Even genuine cancers detected in the thyroid are often small, not aggressive and unlikely to harm the patient, although detection invariably triggers therapies that can permanently damage the nerves and require life-long medication (Welch HG, Schwartz LM, Woloshin S. Overdiagnosed: Making People Sick in Pursuit of Health. Boston, MA: Beacon Press, 2011).
- Gestational diabetes. Around 18 per cent of pregnant women have suddenly found themselves in need of urgent medical treatment after the criterion for gestational diabetes (hypergly-caemia in pregnancy) was altered in 2010, thus doubling the number of ‘sick’ people. Advocates for a more rigorous definition have argued that it reduces health problems at birth, including babies that are ‘large for gestational age’ (Diabetes Care, 2010; 33: 676–82).
- Chronic kidney disease. Around one-third of the over-65s are deemed to have chronic kidney disease and, yet, fewer than one in 1000 of these cases progresses to life-threatening end-stage renal disease every year. The definition for the disease was altered in 2002 (measured as the estimated glomerular filtration rate, or eGFR), a move that has been described as “a fishing trawler that captures many more innocent subjects than it should” (Nephron Clin Pract, 2011; 119 Suppl 1: c2–4). One problem is the eGFR test itself, which is known to be inaccurate. Despite this, elderly people are being diagnosed with kidney disease based on a single measurement, and “the majority of those held to have chronic kidney disease have no identifiable kidney disease” (Nefrologia, 2010; 30: 493–500).
- Asthma. This is a chronic—and sometimes life-threatening—respiratory disorder and, yet, despite its obvious symptoms, more than 30 per cent of patients were so diagnosed when they didn’t have the problem, while around 66 per cent of those who were correctly diagnosed didn’t need the drugs six months later (CMAJ, 2008; 179: 1121–31).
- Pulmonary embolism. A blood clot causing sudden blockage of the main artery of the lung can be fatal, and doctors used to detect it only when there was tissue death (infarction) in the lungs. Today, doctors have the benefit of computed tomog-raphy pulmonary angiography (CTPA), which can detect even the smallest clots. As a result, the incidence of pulmonary embolism has apparently doubled, and has led to “an epidemic of diagnostic testing that has created overdiagnosis” especially of clinically un-important cases that are not life-threatening (Arch Intern Med, 2011; 171: 831–7). While a pulmonary embolism is fright-ening, doctors are so vigilant that they order a CTPA scan at the slightest suspicion; for this reason, the authors call for a degree of common sense.
Seeing more problems
- ADHD (attention-deficit/ hyperactive disorder). This condition has become almost a matter of fashion rather than a genuine diagnosis, and child-ren’s bad behaviour and naughtiness of previous genera-tions have now become a disease that is treated with powerful drugs. To illustrate the point, researchers analyzed the case reports of around one million children in Canada who had been so diagnosed. Those whose birthdays were in December—and thus were among the youngest in their school year—were, on average, 50 per cent more likely to be diagnosed with ADHD than older children (CMAJ, 2012; 184: 755–62).
A change in the definition of what we call disease is certainly one driver of the epidemic of over-diagnosis but, by far, the biggest is the sophisticated screening tech-nology that allows doctors to see an abnormality often before it escalates into disease.
However, the new technology is also discovering something rather interesting: most of us have some ‘abnormality’ that will never develop into a medically recognized disease. This phen-omenon, referred to as a ‘reservoir of abnormalities’, has been high-lighted by researchers who have screened healthy individuals who had no symptoms and nevertheless discovered that:
- around 10 per cent of us have gallstones, as detected by ultrasound, and yet have never displayed any symptoms—such as pain, nausea or problems with fatty foods (Invest Radiol, 1991; 26: 939–45);
- about 40 per cent of us have knee damage, as detected by MRI scans, and yet have no knee pain or even a history of knee injury (Radiology, 1990; 177: 463–5);
- more than 50 per cent of us have bulging lumbar discs, according to MRI scans, but we’ve never experienced any back pain (N Engl J Med, 1994; 331: 69–73).
These figures are averages, and may differ according to age and gender. For example, only 2 per cent of men under the age of 40 have gallstones detected by ultrasound when they have no symptoms compared with 9 per cent of women aged under 40, whereas 80 per cent of both men and women over the age of 50 have bulging discs with no symptoms (Welch HG, Schwartz LM, Woloshin S. Overdiagnosed: Making People Sick in Pursuit of Health. Boston, MA: Beacon Press, 2011).
Another problem is the reliability of the findings. Not only is screening technology picking up symptomless abnormalities, but it also ‘sees’ problems that aren’t even there. Mammography, designed to detect early-stage breast cancer, produces 10 ‘false-positive results’— cancers that aren’t there—for every genuine case it detects (Pol Arch Med Wewn, 2010; 120: 89–94). Indeed, the rise of breast cancer over the last 20 years is almost entirely due to overdiagnosis, or false positives, say researchers from the Norwegian Institute of Public Health. Comparing annual cancer rates for the years 1991 to 2009, new diagnosis levels remained stable in the 40- to 49-year-old group, yet rocketed by 50 per cent in the 50- to 69-year-olds, immediately after routine mass mammography screening was introduced (Tidsskr Nor Laegeforen, 2012; 132: 414–7).Overtreatment
If you do have symptoms and a problem is confirmed by a scan, the next level of medicalization then comes into play—overtreatment.
However, symptoms and a confirmatory test still don’t necessarily point to the correct cause: a patient complaining of knee pain—and whose scan reveals damaged cartilage—is in the fast-track queue for arthroscopic knee surgery. Yet, as we have seen, scans pick up a reservoir of abnor-malities, and it’s possible that the damaged cartilage isn’t the problem and that the pain is, instead, being caused by arthritis. In that instance, knee arthroscopy may well do more harm than good (N Engl J Med, 2008; 359: 1097–107).
The greatest examples of overtreatment are seen when ‘cancer’ is detected, a finding that inevitably sends medicine into overdrive. For prostate cancer, surgery and radiation are the two main assaults—as they are for most cancers. Surgery—called ‘radical prostatectomy’, or complete removal of the prostate and its surrounding tissue—causes life-long sexual dysfunction in half of all patients, and 30 per cent have problems urinating. Radiation has the same side-effects but, in addition, can also damage the rectum; 15 per cent of men treated with radiation develop “a moderate or big problem” with defecation, usually in the form of pain or urgency (J Natl Cancer Inst, 2000; 92: 1582–92; N Engl J Med, 2008; 358: 1250–61).
Women diagnosed with breast cancer face similar, life-debilitating courses of action, including chemotherapy, radiotherapy and/ or breast surgery. The side-effects of chemotherapy drugs include mucositis (inflammation of the mucous membranes), heart and bile-duct problems, bone-tissue death, infertility, and lower white and red blood cell counts.
Many of these cases were misdiagnoses of breast cancer or, in the case of prostate cancer, were not killers. Pathologists discovered the presence of prostate cancer in 80 per cent of men aged between 70 and 79 years who were examined following an accidental death, suggesting that most men develop the cancer, but few die because of it (Cancer, 1989; 63: 381–5).
Given such overdiagnosis and overtreatment, you would at least expect the death rate from cancer to have fallen—but it has hardly altered in the past 30 years. Around 25 men per 100,000 die due to prostate cancer every year and have done so since 1975 and, yet, over the same period, diagnoses have increased from around 90 per 100,000 in 1975 to 150 per 100,000 in 2005, peaking in 1993 with 227 new cases per 100,000 men (SEER Cancer Statistics Review, 1975-2005; http://seer.cancer.gov/ csr/1975_2009_pops09/results_merged/sect_23_prostate.pdf).
A similar picture can be seen for breast cancer. Despite the increase in diagnoses—and subsequent treatment—the number of women dying from the disease has remained the same for 30 years.
In 1975, 27 women per 100,000 died of the cancer; in 2005, the figure was 25 per 100,000. During the same timeframe, new diag-noses were increased by up to 75 per cent, from 100 new diagnoses per 100,000 women in 1975 to 175 per 100,000 in 2001, falling away slightly to 155 per 100,000 by 2005 (SEER Cancer Statistics Review, 1975-2005; http://seer.cancer.gov/csr/1975_ 2009_pops09/results_merged/sect_04_
breast.pdf).Body, heal thyself
Overtreatment is a manifestation of a medical system that sees itself as the guardian against all that ails us, and does not recognize the extraordinary healing powers the body itself possesses.
Stroke, for example, is a debilitating and sometimes fatal disease—and one that has the doctor pushing the panic button. Despite this common perception, many of us have suffered a stroke and have never realized it. One study of around 2000 healthy people discovered that 10 per cent had suffered a stroke, usually a ‘silent stroke’. It also happened to younger individuals: an MRI scan revealed that 7 per cent of the under-50s had had a stroke at some point and, yet, the brain had been able to repair the damage (Stroke, 2008; 39: 2929–35).
Astonishingly, cancer is another feared disease that doesn’t always follow an inevitably aggressive pattern. Breast cancer, for exam-ple, seems to have a six-year cycle, at the end of which it appears to regress. Researchers made the discovery when they analyzed the progress of the cancer in Norwe-gian women who were regularly screened, and compared them with those who were screened only at the beginning and end of the same six years. Although the cumulative rate of cancer was higher in the screened group, as would be expected, those who had cancer detected at the beginning of the six-year period had no trace of the disease when they were screened at the end of the trial (Arch Intern Med, 2008; 168: 2311–6).Think again
Medicalization—the harming of the healthy—is becoming a hot topic in medicine, and politicians and health regulators are starting to join in. In 2010, Archives of Internal Medicine, a prestigious peer-reviewed medical journal, launched a ‘Less is More’ campaign, where doctors were invited to suggest treatments and routine screenings that may be doing more harm than good.
This year, nine physician groups and organizations have joined forces to launch the ‘Choosing Wisely’ campaign, and, as a group, these different specialties have listed 40 procedures or tests that they believe are unnecessary (see box, page 13).
Policy groups across Europe are also discussing ways to tackle excess in medicine, and The Dartmouth Institute for Health Policy and Clinical Practice is holding an important conference next year, entitled ‘Preventing Overdiagnosis’.
Medicine’s own governing bodies are beginning to loosen their definitions of illness, and the first could be a relaxation in the thresholds of high blood pressure. If the proposals go through, 100 million people will suddenly be ‘well’ and will no longer need to take antihypertensive drugs.
In general, overdiagnosis may be the result of a desire to do the best, and to beat disease before it takes hold but, in so doing, medicine has caused enormous damage—and perhaps it is finally waking up to the harm it is doing to the healthy.Bryan HubbardFactfile: 40 procedures you should question (and probably don’t need)
Nine physician groups got together to form the Choosing Wisely campaign (visit www.choosingwisely.org
) to reduce unnecessary procedures and tests. Here are some of their main recommendations. Allergy, asthma and immunology
1. Don’t permit unproven diagnostic tests—such as IgG and IgE—to establish an allergy.
2. Don’t have sinus CT (computed tomography) scans for uncomplicated viral infections, such as rhinosinusitis.
3. Don’t have a routine test if you have chronic urticaria (hives).
4. Don’t have replacement immunoglobulin therapy for recurrent infections.
5. Don’t accept a diagnosis of asthma unless you’ve had a spirometry test. Family health
1. Don’t have imaging screening during the first six weeks of having back pain.
2. Don’t regularly take antibiotics for mild sinusitis (a ‘runny nose’).
3. Say ‘no’ to dual-energy X-rays for osteoporosis screening if you’re female and under age 65, or male and under 70 years of age.
4. Don’t have routine annual screening for cardiac problems if you don’t have any symptoms.
5. Don’t have a Pap smear if you’re under the age of 21 or you’ve had a hysterectomy for a non-cancerous problem.Heart health
1. Don’t have stress cardiac imaging in initial and exploratory testing if you don’t have any symptoms of a heart problem.
2. Don’t have annual stress cardiac imaging as part of routine follow-ups.
3. Don’t have stress cardiac imaging as a preoperative assessment if you’re scheduled for low-risk, non-heart surgery.
4. Don’t have an echocardiogram as a routine follow-up for mild valvular disease.
5. Don’t have stents inserted into the arteries for stable heart attack.General health
1. Don’t have an exercise electrocardiogram if you don’t have any symptoms of heart disease and you’re at low risk.
2. Don’t have imaging screening for low back pain.
3. Don’t have an MRI scan as part of a routine neurological examination.
4. Don’t have imaging screening as the initial diagnostic test if venous thromboembolism is suspected.
5. Don’t have chest radiography if you have no symptoms of a heart problem. Screening
1. Don’t have imaging for uncomplicated headaches.
2. Don’t have imaging for suspected pulmonary embolism as a first-line examination.
3. Avoid a chest X-ray as part of your routine diagnostics before going to hospital.
4. Don’t use CT screening on a child if appendicitis is suspected. Use ultrasound first.
5. Don’t have follow-up imaging for cysts. Gastric health
1. If you have reflux problems, use long-term acid suppression therapy only at the lowest possible dose.
2. Don’t repeat colorectal cancer screening more than once every 10 years.
3. Don’t repeat a colonoscopy for at least five years if you have only one or two small polyps.
4. If you suffer from Barrett’s oesophagus, don’t have a second endoscopy in less than three years.
5. Don’t have a CT scan if you have abdominal pain. Cancer health
1. If you have solid tumours, don’t try therapies that have no track record of success.
2. Don’t have PET (positron emission tomography) or CT scans if you’re in the early phases of prostate cancer.
3. Don’t have a PET or CT scan if you’re in the early stages of breast cancer.
4. Don’t have tests or imaging if you don’t have any symptoms of breast cancer.
5. Say ‘no’ to white-cell-stimulating factors if you have less than a 20-per-cent risk of febrile neutropenia as a complication.Kidney health
1. Don’t have routine cancer screening if you’re on dialysis.
2. If you have chronic kidney disease and a haemoglobin level greater than 10 g/dL without anaemia, don’t use ESAs (erythropoiesis-stimulating agents).
3. Avoid all NSAIDs (non-steroidal anti-inflammatory drugs) if you’re suffering from heart disease, heart failure or chronic kidney disease, including diabetes.
4. If you have acute kidney failure, don’t have central catheters inserted without first getting the permission of your kidney specialist.
5. Don’t have chronic dialysis without getting the agreement of your family and the physicians involved.