Are we diagnosing and treating patients to meet their perceived needs? Studies show that our errors contribute to a trend of overprescribing.
Recently my almost three-year-old grandson had a low-grade fever, and he said his ear hurt. His father was also sick — myalgia and a marginal fever. The morning after, mom took junior to the doctor’s office where the nurse practitioner made a diagnosis.
When I was subsequently advised about the visit, I was virtually positive I knew what the diagnosis was going to be — bilateral otitis media. The prescribed treatment: high-dose (diarrheal dose) amoxicillin (80mg/kg), ibuprofen (he had been on acetaminophen) and advice to take a probiotic.
How unfortunately predictable. Otitis media is a gift from the gods. It creates the perfect opportunity to prescribe what we believe the parents want — antibiotics. Otherwise, why go to the doctor? And does Johnny really have otitis media? I bet a dollar he doesn’t. But otitis media is perfect because the diagnosis cannot be verified by anyone at the time of the visit, and everybody knows you need antibiotics to treat an ear infection.
Convincing Ourselves of the Diagnosis
I believe that in many (most?) cases, clinicians make up the diagnosis of otitis media. They don’t do it fraudulently or with malice, they just convince themselves that that is the diagnosis. Sure, one ear may be a little redder than the other (studies have demonstrated that red ears are most indicative of viral infections, not bacterial ones). And do we check to see if there is fluid behind the TM? There are tools to verify if there is or is not, but virtually nobody uses them in the ED. (We prefer to use the “art” of medicine to make the diagnosis we want to make.)
The first paper we’ll review this month is 36 years old, from back when bacterial otitis media was much more common than it is now. This was before pneumococcal and H. influenza immunizations (the two most common causes of bacterial otitis media). Even then, as the first study indicates, clinicians were making the diagnosis pretty much on whatever criteria they wanted. Why? Because clinicians felt more comfortable giving antibiotics (“just in case something more serious was causing the fever”) and parents came to expect them. This was the age when we were “sheep dipping” children in antibiotics.
More specifically, the study notes that in a 25-year literature review (that would take the data back 61 years) and in a study of 185 pediatricians, 18 separate sets of criteria were used to make the diagnosis in the clinical studies, while the pediatricians proposed 147 different sets of criteria. Pretty impressive. With a choice of this many criteria, clinicians can make the diagnosis just about whenever they want (or need) to.
ACUTE SUPPURATIVE OTITIS MEDIA IN CHILDREN DIVERSITY OF CLINICAL DIAGNOSTIC CRITERIA
Hayden, G.F., Clin Pediatr 20(2):99, February 1981
The author, from the University of Virginia Medical Center, conducted a 25-year literature review (43 studies) and surveyed 185 pediatricians to determine suitable criteria on which to base a clinical diagnosis of acute otitis media. Only 60% of the studies reviewed (26) described diagnostic clinical criteria, and in these 26 studies 18 separate sets of criteria were utilized. Tympanocentesis was used to confirm the diagnosis in eleven of the studies, although these did not describe clinical diagnostic criteria, and the indications for performing this procedure were not set forth. The survey of pediatricians revealed that a diagnosis of acute otitis media is made in approximately 13% of patients. Generally it was felt that visualization of 70% of the tympanic membrane was required for an adequate ear exam, and some pediatricians stated that certain areas, such as the bony landmarks, light reflex, and pars flaccide, were particularly important to diagnosis. In 24% of cases, removal of cerumen (preferably with curettage and flushing with a syringe or Water-Pik) was necessary for adequate exam. A total of 43% of participants regularly used a pneumatic otoscope. In 77% of cases, the pediatricians felt that their diagnosis was definitely accurate, and that it was “probable” in an additional 17%. The 165 respondents (89%) who listed their specific clinical criteria for the diagnosis of acute otitis media proposed 147 different sets of criteria. Through use of the pediatrician questionnaire, a ranking order of 20 diagnostic signs was established. The signs felt to most reliably indicate otitis media were: bulging TM, otorrhea, red TM, complaint of ear ache in an older child, and absent/distorted boney landmarks. It is concluded that there is a tremendous diversity of practice and opinion regarding the clinical diagnosis of acute otitis media. Further clinical studies corroborating physical finding with objective measure of infection (tympanometry and tympanocentesis) are needed. 31 references 8/81-#16
So junior got his prescription of amoxicillin filled (and was advised to take half the dose by grandpa — a modest 40mg/kg), and the next morning he was not complaining of any ear pain after being given just one miraculous dose. One dose cures you! And he went to nursery school today and swimming lessons. Ah, modern medicine. My contention: The child did not have bilateral otitis media, at least not the suppurative kind, and if he did have BOM, it almost assuredly was viral and that he met criteria for observation, according to the 2013 criteria of the American Academy of Pediatrics.
But what about a diagnosis where lots of people can potentially assess the site of pathology: The sore throat? It’s going to be a lot harder to ascribe a bacterial infection here as the cause, a place where objective evidence is more apparent — submandibular adenitis, exudate, fever, absent URI symptoms. Well, no problem. We still can make a diagnosis of bacterial sore throat and give out that amoxicillin any time we want.
Warning: Read the Entire Abstract
The next paper is one of my all-time absolute favorites. Perfectly well actors (who had their throats photographed to document the absence of any pathology) told doctors one of two stories, one consistent with a viral infection and another consistent with a bacterial infection. The results were most disheartening, but consistent with the otitis study above. So as to not short-charge the reader, you really have to read the whole abstract to appreciate just how badly and deceptively the physicians performed.
BIAS IN THE EVALUATION OF PHARYNGITIS AND ANTIBIOTIC OVERUSE
Kiderman, A., et al, Arch Intern Med 169(5):524, March 9, 2009
METHODS: These Israeli authors evaluated the influence of bias introduced in a patient history on physicians’ perceptions regarding clinical findings and actual management. Healthy actors visited 32 clinicians (30 trained outside the U.S.), reporting a history consistent with viral infection (headache, fever, cough and runny nose for two days with throat discomfort and hoarseness on the day of the visit) or bacterial infection (sore throat for one day with headache and fever with malodor of the mouth but without cough or nasal discharge). None of the actors had physical findings consistent with illness, as confirmed on pre-visit evaluations and photography.
RESULTS: The experience level of the participating physicians ranged from 5 to 32 years (mean, 19 years), and 13 of the physicians were board-certified in family medicine. The physicians recorded slight, moderate or severe pharyngeal erythema for 41%, 34% and 6% of the actors presenting the viral script, and for 22%, 31% and 22%, respectively, of those presenting the bacterial script. An exudate was recorded for 6% and 25% of the actors presenting the viral and bacterial scripts, respectively, and lymphadenopathy was recorded for 16% and 26%, respectively. Throat culture was done for 47% of the actors presenting the script consistent with viral illness, and for 73% of those presenting the bacterial illness script, and antibiotics were prescribed for 21% and 79%, respectively.
CONCLUSIONS: These findings demonstrate that physicians often “find” physical findings consistent with what they expect to find, based on a patient’s history, and that this appears to be true regardless of the level of physician experience.
5 references (email@example.com – no reprints). Copyright 2009 by Emergency Medical Abstracts – All Rights Reserved 7/09 – #14
So what’s the point of all this? The studies show the extent clinicians will go to meet the perceived expectations of patients with regard to testing and treatment. They also demonstrate, at least in the case of the latter study, how poorly this group of physicians followed evidence-based approaches to the patient with a history of a sore throat. The drive to err on the side of treatment is powerful, used to satisfy patients by giving them what we think they want (whether it is truly what they want or not). So it seems that we need to reinforce the obligation to “keep up” in medicine. We are life-long learners. We cannot practice as we did when we finished out residencies 20 years ago. And the medical community needs to embrace the fact that overtesting and overtreatment is rampant, and that can be harmful and unethical.