Listen to your patient; he is telling you the diagnosis. —Dr. William Osler
I am not aware that I have ever had a mistaken diagnosis. —Child Abuse Pediatrician Dr. Barbara Knox
Doctors are human. That should be obvious, but people seem to expect doctors to be perfect the first time, every time. Patients notice (and sometimes sue) when a doctor falls short.
Despite their best efforts, doctors make mistakes individually and corporately. Because they know so much more about the human body than do non-doctors, there can be a tendency to place too much trust in their opinions, even to interpret their opinions as facts.
Medical mistakes and misdiagnoses occur more often than most people realize. In Improving Diagnosis in Health Care, diagnostic error is defined as “failure to establish an accurate and timely explanation of a patient’s health problem or to communicate that explanation to the patient.”
A study published in January of 2024 in JAMA Internal Medicine states that nearly 1 in 4 hospital patients who died or were transferred to intensive care had been misdiagnosed. A 2024 study published in BMJ estimates that 795,000 patients a year die or are permanently disabled due to misdiagnosis or delayed diagnosis.
Medical errors are the third most common cause of death in the United States. Only heart disease and cancer claim more lives. 251,000 annual deaths due to misdiagnosis equals about 700 deaths a DAY! Autopsy studies reveal that an estimated 15-25% of patients who die had been misdiagnosed.
The problem of diagnosis is confounded when:
- The diagnosis itself is unreliable, as in the case of Shaken Baby Syndrome.
- Doctors believe one diagnosis is far most “likely” than other possibilities.
- A doctor places too much confidence in first impression.
- A doctor distrusts the history given by a caretaker or eyewitness. Once there is a suspicion of abuse, mistrust quickly distorts everything.
- The patient is a baby or a child who cannot clearly communicate his/her symptoms.
- A patient dies before the correct diagnosis is found.
- Emotion overwhelms reason. Grieving parents and a dying baby tax one’s ability to remain detached and rational.
- A rare disease is involved. More than 7,000 rare diseases exist that an average doctor never encounters. Rare diseases may not even be considered when abuse is suspected.
Doctors may be pressured to provide the reason for a baby’s suffering immediately if not sooner. “I don’t know” is not satisfactory, even when true. Suspicions can quickly morph into “facts” once they’re stated or written down.
An expert clinician typically forms an idea of what’s wrong with a patient within 20 seconds. Once a diagnosis is made, it becomes “sticky”. Not only the diagnosing doctor, but specialists who review the chart find it difficult to consider alternatives or think outside the box. It is not unusual for a doctor to become wedded to a mistaken diagnosis and develop an emotional and psychological commitment to it.
The process of differential diagnosis requires doctors to consider alternative explanations for a clinical picture. They list several possibilities and rank them in order of likelihood. Tests are then ordered to rule out, to the degree possible, some of the alternatives. After weighing costs and benefits, the doctor decides on a plan. It is not worthwhile, for example, to order an expensive or painful test to rule out a very rare disease. Sometimes the best way to test a diagnosis is to treat the patient for it, especially if possible side effects are minimal. If the patient improves, the diagnosis is confirmed. If not, the doctor considers another diagnosis and works down through the list in order of likelihood until the true diagnosis is found.
Differential diagnosis works well in most cases, but not with Shaken Baby Syndrome. Aside from the fact that SBS assumes etiology, there is no way to decisively “rule out” abuse. If a doctor thinks that a caretaker is lying because the baby’s condition doesn’t appear “consistent with” expectations, suspicion will color everything about the case from that point forward.
People tend to see what they look for. Child Abuse Pediatricians look for child abuse. They claim to rule out abuse in 40% of cases referred to them, but when a child dies or is seriously harmed, a quick diagnosis (or misdiagnosis) can be made with tremendous confidence.
One study compared 20 radiologists having a diagnostic accuracy of nearly 95% with 20 having a diagnostic accuracy of just 75%. Surprisingly, the radiologists who performed poorly were more confident in the accuracy of their readings than were the top 20. Overconfidence contributes to diagnostic error in medicine. Charles Darwin wrote, “Ignorance more frequently begets confidence than does knowledge.”
For years it was taught that retinal hemorrhages were “pathognomonic” of violent shaking. Pathognomonic means uniquely associated with or diagnostic of. A doctor who suspected abuse would call an ophthalmologist. If retinal hemorrhages were found, the diagnosis of Shaken Baby Syndrome was confirmed. It was elegantly simple and straightforward. No one looked into the eyes of “normal” children to establish baselines.
While it may be acceptable to misdiagnose a cause of death, things are different when the diagnosis is murder and criminal charges are lodged. Every known alternative should be ruled and it should be kept it mind that the death might be due to a cause as yet unknown. Even if 95% of children with the symptoms attributed to Shaken Baby Syndrome were truly abused, what about the other 5%? Is it acceptable to presume criminality?
Every doctor since the 1980s has been trained to put Shaken Baby Syndrome at the TOP of a differential diagnosis list when retinal and subdural hemorrhages are found in a baby. As soon as medical findings suggest SBS, doctors disregard the history provided and substitute a presumed history of violent shaking.
Doctors are not “bad” because they err. They are human. Most do the very best they can to care for patients without all puzzle pieces being on the table.
The problem with SBS goes back to how doctors were trained and what they were taught. It’s unrealistic to expect doctors to be perfect and it’s unrealistic to expect them to critically evaluate the scientific basis for what they were taught. They reasonably assume someone else did that before it appeared in medical school textbooks!
Here is a list of a few confirmed mistaken diagnoses of Shaken Baby Syndrome or child abuse. There are many more. No one knows how many misdiagnoses have not been confirmed. In most of the cases below, charges were dropped. In many, Child Abuse Pediatricians maintain that despite the proven alternate diagnosis, the child was still abused. “Just because a baby has a congenital condition doesn’t mean she wasn’t shaken,” they say. That’s true, but if a congenital condition fully explains the medical findings, why surmise criminality?
2018 Tomball, Texas — The two children of a stay-at-home mom were playing in the sprinkler on a hot day. While drying her daughter, the mother sat her son on a chair. He fell and was raced to the hospital. Both children were removed from their home. Months later, a judge cleared both parents after a CPS worker lied in court and CPS was ordered to pay the parents’ legal fees. This case led to a law in Texas allowing parents to seek a second opinion.
2008 Commerce City, Colorado — Just prior to criminal charges being filed, married police officers were found dead in their home after an apparent murder-suicide. They had been accused of abusing their 3-month-old daughter, who was forcibly removed from their home by authorities. The baby was later found to have spinal muscular atrophy, which fully accounted for the medical findings initially attributed to abuse.
2007 Florissant, Missouri — A baby sitter who had operated a home daycare for years found an 11-month-old boy unresponsive. He was rushed to the hospital and found to have retinal hemorrhages and brain damage. He subsequently died and second-degree murder charges were filed. Luckily, another couple whose child had similar symptoms read the story in the newspaper and called the sitter’s family to suggest that the baby be tested for glutaric aciduria. A test was done and the diagnosis was confirmed, clearing the sitter of blame.
1991 Denver, Colorado — A 7-month-old girl was brought to Colorado Children’s Hospital after she fell and hit her head. A team of physicians diagnosed abuse. The baby was removed from home and the parents were criminally charged. Luckily, someone eventually tested the baby for a rare disorder, glutaric aciduria type 1 (GA-1). The test was positive. In the end the parents won a lawsuit against the hospital and physicians. In part because the correct diagnosis was delayed, the child became permanently disabled. Had she died, the correct diagnosis would never have been found and the parents may have been convicted of murder.
1993 Fort Campbell, Kentucky — The baby fell off the bed during a diaper change and was taken to the hospital. When physicians at Vanderbilt University Medical Center found retinal hemorrhages, the mother, an Army physician, was accused. The baby was placed in a foster home and a 19-month-long custody battle began. The mother combed through the medical literature searching for an explanation. Eventually, she abducted the baby and took him to specialists in another state, where he was diagnosed with Alagille’s Syndrome. A judge reunited the family after hearing testimony from defense experts.
1997 Charles City, Iowa — The mother of a 4-month-old boy was charged with first-degree murder after a coroner ruled the death a homicide. On the third day of jury selection, the State moved to dismiss the charges after convincing evidence from defense experts proved the baby died of an aneurysm.
1997 El Cajon, California — Retinal hemorrhages in a 2-month-old boy were found during a routine visit to the pediatrician. Although the baby had been born with Rubenstein-Taybi Syndrome, a rare disorder, the parents were charged with child abuse and the baby was removed from the home. It was a year before charges were dropped. The parents filed an unsuccessful civil suit against the county.
1998 Vandalia, Ohio — At 2 months of age, a baby boy was diagnosed with Shaken Baby Syndrome. He and his siblings were removed from the home and adopted out, though no criminal charges were filed. The baby was severely handicapped and died several years later. The death was ruled a homicide. A year after his death, a full-blooded brother was born. He died at 6 months of age and was correctly diagnosed with Menke’s Disease. The body of the first child was exhumed and it was confirmed that he suffered from the same genetic disease. The mother, however, was unable to reunite with her living children. They had been permanently adopted into other families.
1998 Vero Beach, Florida — Both children were removed from the home after the mother was accused of shaking her 1-month-old daughter. The children were not allowed home for 18 months. Charges were finally dropped after doctors proved the baby’s condition was due to a birth injury.
Considering all of this, which of these scenarios seems more likely?
1. A person with a long, excellent record of caring for children one day “loses it” and violently shakes a baby, expresses no remorse, then lies by claiming not to have done so and continues to lie after conviction when a refusal to acknowledge guilt means no parole.
2. Doctors don’t know everything and can make mistakes.