In this cost-cutting age, clinicians can’t take every step to confirm a diagnosis. But if you’re wrong, you’re in peril.
Dr. Y, age 47, is an experienced family physician who prides himself on “effective management” techniques. He believes that efficient diagnostic algorithms combined with testing and trial therapies can produce an accurate diagnosis at a reduced cost, allowing treatment to start as quickly as possible.
A case of suspected carpal tunnel syndrome led to a courtroom test of his theories.
The patient was a 38-year-old bank teller who noticed increasing numbness in her hands over two months. The pain had initially responded to aspirin, which she took regularly, but over the past week, it had become worse. She wanted something stronger.
She blamed her pain on the register she used to track deposits and disbursements. It was an older mechanical model, she said, that was harder to use than the modern “soft-key” machines available to other tellers. After comparing notes with her colleagues, she believed “carpal tunnel” accounted for her symptoms.
Dr. Y examined the patient’s hands and found the symptoms seemed to be more related to pain and stiffness in the wrists than the typical carpal tunnel case of median nerve compression and resulting numbness. Based on a working diagnosis of wrist arthritis, he began treatment with nonsteroidal anti-inflammatory drugs (NSAIDs). His thought process here, as he later had to explain to the jury, was to avoid the expense of exhaustive testing, which would have cost the patient hundreds of dollars.
This “trial therapy” approach had worked well for Dr. Y over the years, since he was usually correct in his provisional diagnosis. He did, however, take the precaution of ordering liver function tests and a complete blood count because the patient had been taking aspirin, which can cause gastric bleeding, and he was giving her prescription-strength NSAIDs, which can exacerbate liver disease. He then discharged the patient for follow-up in one week.
When the patient’s test results came back, he noted a hemoglobin level of 11.0 g/dL and no abnormalities in the liver profile. He attributed the anemia to the patient’s aspirin use and to possible blood loss associated with asymptomatic gastritis and esophagitis, which he knew was common. At her next visit, the patient reported relief of her wrist pain from the NSAID, and Dr. Y reassured her about the results of her liver tests and blood work.
Dr. Y continued to follow the woman over the next nine months, during which time her symptoms gradually got worse. Even NSAIDs did not fully relieve her pain anymore. Dr. Y also noted her joints had become enlarged and swollen. After discussing other diagnostic possibilities with the patient, he referred her to a joint specialist, who ran several thousand dollars’ worth of tests before concluding that she had rheumatoid arthritis (RA) affecting her wrist and polyarteritis nodosa affecting her liver and blood count. He started her on steroids and anti-RA drugs, with good effect.
The patient stopped seeing Dr. Y while remaining under the specialist’s care. Dr. Y lost track of her but heard that she did well for several years. Then she developed liver disease that required transplantation. Six months later, Dr. Y received notice that the patient was suing him for misdiagnosis.
The case progressed to depositions. The patient appeared first and testified that she had described pain and swelling in both wrists at her first visit with Dr. Y. But the physician had ignored the possible diagnosis of RA and presumed the more common condition of nonspecific arthritis. She then gave a graphic account of the damage that her illness had caused, culminating in the liver transplant and all the morbidity and pain that can accompany that procedure.
Dr. Y gave his deposition. In a visit-by-visit analysis of the chart, Dr. Y explained his reasoning and management at each stage of the disease, taking the chart entries and expanding on each to create a coherent account of the case.
After the depositions were completed, attempts to settle the case failed because the two sides assessed the value of the case very differently. While the plaintiff’s lawyer assessed the value at $2 million because of the patient’s sympathetic predicament, the defense lawyer assessed it at $50,000. This amount was deemed a “nuisance value” settlement and was calculated to cover the plaintiff’s lawyer’s expenses.
The case went to trial about two years after it was originally filed, and following live testimony from the patient and Dr. Y, as well as videotaped depositions from the experts on both sides, the jury held in favor of Dr. Y.
Most cases settle, and just as well for the judicial system, which is creaking under the weight of the great American pastime: litigation. But for a case to settle, the plaintiff’s lawyer and the defense team must have a similar view of the settlement value. This is different from believing in the rightness of the cause on each side but rather is an approximation of how much the case would bring before an average jury in that region, reduced by a factor related to the chance of winning the case. Many disputes are settled through mediation, which can effectively short-circuit the lengthy and expensive process of trial in the courtroom.
According to the figures of most malpractice insurance companies, misdiagnosis is the most common ground for a suit. Researchers are studying the ad hoc rules and processes by which clinicians make diagnostic decisions (known as medical “heuristics”) and are developing improved diagnostic algorithms.
In this case, Dr. Y assumed the more common diagnosis of osteoarthritis of the wrist rather than the less common—but in this patient more accurate—diagnosis of RA. The situation illustrates the clinician’s dilemma: how much testing to perform to confirm a common and probable diagnosis. Although this point of balance varies with the clinician, from a risk-management point of view, it poses a problem, since one can imagine the plaintiff’s expert testifying: “If only Dr. X had performed this one simple test, the case would have turned out differently.”
To anticipate and counter this situation, clinicians should enter enough detail in the medical chart so they can justify why they stopped the investigative process when they did. A short notation, such as “prov diag OA bilat, trt and review 1 wk,” is sufficient. When interpreted appropriately by the defending clinician, it explains the approach taken in the case.