Bedside Rounds

Bedside Rounds: How Useful are the Kernig and Brudzinski signs for Predicting Meningitis?

April 27, 2016

kernigBy Chio Yokose, MD

Peer Reviewed

Even in this era of modern medicine, bacterial meningitis remains a widely feared diagnosis in both resource-rich and -poor settings worldwide. Bacterial meningitis is among the ten most common infectious causes of death and kills approximately 135,000 people around the world each year [1].

It is a medical, neurologic, and sometimes neurosurgical emergency that affects 4 to 6 per 100,000 adults annually [2].  Many healthcare providers may consider the diagnosis when evaluating a patient, but it can nonetheless be difficult to recognize and act on. Any such delay can be the difference between life and death.  According to a retrospective study, the median delay between the time of arrival at the emergency department and the administration of antibiotics was reportedly 4 hours [2].

A lumbar puncture is the diagnostic test of choice for meningitis. However, lumbar punctures are known to be somewhat invasive and unpleasant tests that have their own set of complications. How do we identify those patients who truly need one? For this purpose, we resort to the timeless history and physical. However, clinical history alone is not sufficient to arrive at a diagnosis of meningitis. According to a systematic review by Attia et al, symptoms of nonpulsatile headache, generalized headache, and nausea or vomiting offer only 15%, 50%, and 60% specificities for meningitis, respectively [3]. As far as the physical exam goes, the classic triad associated with meningitis is fever, neck stiffness, and change in mental status. According to Attia et al’s systematic review, although few patients presented with all three symptoms, 95% demonstrated two or more and 99-100% had at least one of these findings [3]. Van den Beek et al reported similar findings from their study of 1108 cases of meningitis identified in the Netherlands Reference Laboratory for Bacterial Meningitis database. They found that, while only 44% of patients demonstrated all three features of the triad, 95% of patients presented with at least two of the four signs (classic triad plus headache) and only 1% of cases had none of the four findings [4].

The singular most common finding associated with meningitis is uncertain, as there is varying data from different observational studies. According to a single-center, multi-decade study, fever appeared to be the most common, with 95% of patients presenting with fever and another 4% developing a fever within the first 24 hours of hospitalization [5]. In contrast, Van den Beek’s Netherlands study found that headache was the most common finding, occurring in 87% of patients [4]. Both studies found that neck stiffness was the second most common finding seen in 88% and 83% of patients, respectively.

No physical exam for meningitis is complete without mentioning the Kernig and Brudzinski signs. Although nuchal rigidity, one of the hallmark features of bacterial meningitis was recognized as early as the 5th century BCE, these major eponymic signs – still so closely linked to the disease to this day – were not described until the late 19th century [6].

Vladimir Mikhailovich Kernig (1840-1917) was a clinical neurologist of Russian-Baltic German descent who was born in Lapaia, Latvia but received the majority of his professional training in Russia [7].  In 1882, Kernig described the sign that now bears his name, to a group of his colleagues in St. Petersburg:

“I have observed for a number of years in cases of meningitis a symptom which is apparently rarely recognized although it is, in my opinion, of significant practical value. I am referring to flexion contracture of the legs or occasionally also in the arms which becomes evident only after the patient sits up…If [with the patient sitting on the edge of the bed and legs dangling] one attempts to extend the patient’s knees one will succeed only to an angle of approximately 135 degrees.  In cases in which this phenomenon is pronounced, the angle may even remain at 90 degrees [6].”

Today, this maneuver is performed with the patient in the supine position with the hips and knees in flexion. The knee is then slowly extended.  The Kernig sign is said to be positive if this maneuver elicits pain along the hamstring muscle as a result of stretching of the inflamed sciatic nerve [7].

BrudzińskiJosef Brudzinski (1874-1917) was a Polish-born pediatrician who also received most of his training in Russia [7].  He actually described several different physical signs of meningitis (e.g. the cheek sign and symphyseal sign, both described further below, and Brudzinski contralateral reflex, which consists of reflex flexion of a lower extremity in response to passive flexion of the contralateral lower extremity [8]), however his most famous sign that is now referred to as the “Brudzinski sign” was described as follows in 1909:

“I have noted a new sign in cases of meningitis: passive flexion of the neck causes the lower extremities to flex at the knees and the pelvis…With the child in the supine position, the examiner flexes the neck of the child with the left hand while resting his right hand on the patient’s chest to prevent it from rising [6].”

A patient is said to have a positive Brudzinski sign if the passive flexion of the neck elicits automatic flexion at the hips and knees [9].  Interestingly, the cheek sign, which is considered positive if applying pressure on both cheeks inferior to the zygomatic arch causes a spontaneous flexion of the forearm and arm [10], and symphyseal sign, which is considered positive if pressure applied to the pubic symphysis elicits a reflex hip and knee flexion and abduction of the leg, were most commonly observed in children with Mycobacterium tuberculosis meningitis, which was much more prevalent during Brudzinski’s time [7].

The utility of using the Kernig and Brudzinski signs to identify patients likely to have meningitis has long been debated. Both of these signs are indicators of meningeal inflammation, but neither are pathognomonic for meningitis [9].  Interestingly enough, Brudzinski himself published a study in 1909 reporting the sensitivities of Brudzinski and Kernig signs as 97% and 42% respectively [11].  These sensitivities are higher than those reported in more recent studies.  Some attribute this difference to the fact that the two most common causes of meningitis in Brudzinski’s time were Streptococcus pneumoniae and M. tuberculosis, both of which are known to cause greater degrees of meningeal inflammation than other infectious etiologies that are more prevalent today [7].

A study by Durand et al conducted at Massachusetts General Hospital between the years 1962-1988 demonstrated that, while S. pneumoniae was the most common pathogen identified in cases of community acquired meningitis overall (responsible for 24% observed 493 episodes), it was not an overwhelming majority [5]. In this study, gram-negative bacilli other than Haemophilus influenzae were responsible for 17% of the cases, whereas N. meningitides, other streptococci, Staphylococcus aureus, and Listeria monocytogenes caused 7-8% each. H. influenzae was identified as the causative organism in only 4% of the cases, and in 15% no pathogen was ever identified. The relative frequency of S. pneumoniae waned significantly in the period 1980-1988 compared to an earlier decade, 1962-1970. In this study, tuberculous meningitis was not identified in a single community or nosocomial case. Today, S. pneumoniae meningitis carries a poor prognosis independent of other factors, exhibiting odds of an unfavorable outcome six times as higher than does N. meningitidis (95% confidence interval, 2.61-13.91; P<0.001) [4].

More recently, Thomas et al published a study in 2002 examining the diagnostic accuracy of Kernig and Brudzinski signs and nuchal rigidity, in adult patients with suspected meningitis [12].  Adults (age > 16 years) presenting to Yale-New Haven Hospital Emergency Department between July 1995 and June 1999 with clinically suspected meningitis were eligible for participation. The study found that both Brudzinski and Kernig signs had a sensitivity of 5% with a positive likelihood ratio (LR+) of 0.97. The study concluded that these special physical exam findings did not accurately discriminate between patients with meningitis (defined as having 6 x 109/L white blood cells (WBCs) per milliliter of CSF or more) and those without. The diagnostic accuracy of these signs was not significantly improved in patients with moderate meningeal inflammation (defined as 100 x 109/L WBCs per milliliter of CSF or more).

Similar results were published by Waghdhare et al in a blinded study of 190 patients diagnosed with meningitis at a rural teaching hospital [1]. Kernig sign had a reported sensitivity of 14.1%, specificity of 92.3%, LR+ of 1.84, and a negative likelihood ratio (LR-) of 0.93. Brudzinski sign had a reported sensitivity of 11.1%, specificity of 93.4%, LR+ of 1.69, and a LR- of 0.95. This study also examined the head jolt sign, which is considered positive if there is a worsening of the baseline headache when patients are asked to turn their heads horizontally at a frequency of 2-3 rotations per second. The head jolt sign had a sensitivity of 6.1%, specificity of 98.9%, LR+ of 5.52, and a LR- of 0.95. In this study, the Kernig sign was positive in only 12% of 190 patients. Thus, although all of these signs have high specificities, the positive predictive values remain low, suggesting they are of little clinical utility in appropriately identifying patients that warrant further diagnostic work-up or treatment of meningitis.

Of note, the age of the patient in question may impact the utilities of these physical exam tests. Interestingly, the majority of patients examined in Kernig’s and Brudzinski’s original papers were children.  Ironically, it has become widely accepted that neither the Kernig or Brudzinski signs are reliable in diagnosing meningitis in infants younger than six months [7].  A retrospective study conducted by Levy et al demonstrated that, as the age of the patients increased (from 2-24 months to 5-12 years), the sensitivity of both the Kernig and Brudzinski signs also increased [13].  A similar trend is noted at the other end of the age spectrum, with Puxty et al reporting that Kernig sign was positive in 12% and Brudzinski sign in 8% of elderly patients on general medicine wards without bacterial meningitis, which is hypothesized to be related to the increasing incidence of cervical spine pathology in this age group which can complicate the interpretation of these maneuvers [14].

Thus, although we continue to regularly test for and document the Kernig and Brudzinski signs in our physical exams regularly, the current data suggest that it may be more of a historical finding and not one that is particularly helpful in delineating a case of bacterial meningitis based on the presence or absence of these signs alone.

Dr. Chio Yokose is a 3rd year resident at NYU Langone Medical Center

Peer reviewed by Michael Janijigian, MD, internal medicine, NYU Langone Medical Center

Image courtesy of Wikimedia Commons


  1. Waghdhare, S., et al., Accuracy of physical signs for detecting meningitis: a hospital-based diagnostic accuracy study. Clin Neurol Neurosurg, 2010. 112(9): p. 752-7.
  2. van de Beek, D., et al., Community-acquired bacterial meningitis in adults. N Engl J Med, 2006. 354(1): p. 44-53.
  3. Attia, J., et al., The rational clinical examination. Does this adult patient have acute meningitis? JAMA, 1999. 282(2): p. 175-81.
  4. van de Beek, D., et al., Clinical features and prognostic factors in adults with bacterial meningitis. N Engl J Med, 2004. 351(18): p. 1849-59.
  5. Durand, M.L., et al., Acute bacterial meningitis in adults. A review of 493 episodes. N Engl J Med, 1993. 328(1): p. 21-8.
  6. Tyler, K.L., Chapter 28: a history of bacterial meningitis. Handb Clin Neurol, 2010. 95: p. 417-33.
  7. Ward, M.A., et al., Josef Brudzinski and Vladimir Mikhailovich Kernig: signs for diagnosing meningitis. Clin Med Res, 2010. 8(1): p. 13-7.
  8. Mehndiratta, M., et al., Appraisal of Kernig’s and Brudzinski’s sign in meningitis. Ann Indian Acad Neurol, 2012. 15(4): p. 287-8.
  9. Wartenberg, R., The signs of Brudzinski and of Kernig. J Pediatr, 1950. 37(4): p. 679-84.
  10. Verghese, A. and G. Gallemore, Kernig’s and Brudzinski’s signs revisited. Rev Infect Dis, 1987. 9(6): p. 1187-92.
  11. Brody, I.A. and R.H. Wilkins, The signs of Kernig and Brudzinski. Arch Neurol, 1969. 21(2): p. 215-8.
  12. Thomas, K.E., et al., The diagnostic accuracy of Kernig’s sign, Brudzinski’s sign, and nuchal rigidity in adults with suspected meningitis. Clin Infect Dis, 2002. 35(1): p. 46-52.
  13. Levy, M., E. Wong, and D. Fried, Diseases that mimic meningitis. Analysis of 650 lumbar punctures. Clin Pediatr (Phila), 1990. 29(5): p. 254-5, 258-61.
  14. Puxty, J.A., R.A. Fox, and M.A. Horan, The frequency of physical signs usually attributed to meningeal irritation in elderly patients. J Am Geriatr Soc, 1983. 31(10): p. 590-2.


Lies My Patients Told Me: “I Take My Medications Every Day.”

January 15, 2016

BloodPressure2By Rebecca Sussman

Peer Reviewed

Reviewing medical evidence has become such a habit that sometimes it feels almost impossible to think independently. I’ve always been a top-down thinker; I go with my gut instinct, and then look for the evidence to support my assessment.

The problem is that very often it feels like what patients need most is not the precision of a particular etiology or the selection of a medication that is perfectly and precisely tailored to their condition and comorbidities; what they need is education about what it means to maintain their health, and practical strategies for how to do so. And my long hours delving into the literature for evidence on how best to do that have been less helpful than I’d hoped.

Says the patient, “I take my medications every day.” Literature on this subject is, in fact, adequately robust for me to mistrust those words when uttered by the majority of patients. For example, toxicological monitoring (which is pretty sensitive, if not practical in the day-to-day clinic setting) reveals that a whopping 50-60% of patients with resistant hypertension are non-adherent with their medications [1].  A cohort study done in Quebec and published in 2014 found that 31.3% of 37,506 first-time prescriptions were not even filled, much less taken [2].

I blame myself for the patient’s fib. If my patients feel that they cannot be honest with me about missed doses, that’s a reflection of my own failing to develop rapport. One of the best pieces of advice I have received in medical school is to start with, “I know that when I need to take medications it’s very hard for me to remember every dose. How many pills do you think you’ve missed this week?” I’ve put it to good use, but where is the literature on how to ask questions that will yield honest answers? It may be there, but I’ve yet to find it, outside of maybe the most psychodynamically-oriented mental health journals.

There is, however, a growing body of research into validated tools to assist in screening for medication adherence. My personal favorite is the brand-spanking-new Measure of Drug Self-Management, abbreviated as MeDS—cute, right [3]? But it’s not the catchy name that draws me to this newly-developed and validated 12-item questionnaire; rather, it’s that the authors specifically sought to develop an inexpensive tool that is considerate of the wide range of patient behaviors (and barriers) and that applies to a diverse range of patients with variable literacy levels [3]:

  1. Did you forget to take your (insert drug name) at any time last week?
  2. In the past month have you stopped taking (insert drug name) for any reason without telling your doctor?
  3. I often forget to take my medicine.
  4. I am organized about when and how I take my medicines.
  5. I have a hard time paying for my medicines.
  6. The print instructions of my prescription bottles are confusing.
  7. Having to take medicines worries me.
  8. I often have a hard time remembering if I have already taken my medicine.
  9. I do not take my medicines when I am feeling sad or upset.
  10. My medicines disrupt my life.
  11. When my medicine causes minor side effects, I stop taking it.
  12. The idea of taking medications for the rest of my life makes me very uncomfortable.

The tool is practical, efficient, and patient-centered–all of the things I strive to be. However, as it has only been publicized within the past month, I’d like to see some more validity testing before investing in the MeDS scale myself. Part of the reason for my skepticism is a lack of evidence-based interventions that physicians can recommend to help patients improve their medication adherence. A Cochrane review published in 2014 analyzed a total of 182 randomized controlled trials aimed at enhancing medication adherence for a wide range of patient populations and medical conditions, and concluded that the methods used for researching such interventions were insufficiently advanced [4]. In essence, no conclusions can be drawn on the basis of existing research because we are not sufficiently adept at performing this research in the first place.At what point, then, do I give up on finding the evidence to back up my instincts? Do I withhold my suggestions from patients until I know that we’re on the right track? Do I start doing the research myself? How could my methods possibly be more advanced than those of the Cochrane review? It brings me back to how I framed the issue for a patient who was frequently skipping breakfast: You should try to get more of your calories in earlier in the day. Theres research to support that. Ive heard the explanation that when you eat in the morning, thats fuel for your body and you burn those calories during your daily activities. But the calories that you eat before bed go right into storage, because youre just going to bed and not doing anything active. Thats just anecdotal, though—I’m not sure what science says about that. It kind of makes sense though, right? Im just throwing it out there to help you feel motivated and empowered to change the way you eat, so that you dont follow in your dads footsteps of having a heart attack at age 50.So I printed out the Mayo Clinic page on Mediterranean diet and sent the patient home to do some reading about plant-based diets. I’m still, however, flooded with a sense of inadequacy and powerlessness when it comes to educating myself on how to foster the trust of my patients and educate them appropriately.

Rebecca Sussman is a 3rd year medical student at NYU School of Medicine

Peer reviewed by Michael Tanner, Associate Professor of Medicine, Executive Editor, Clinical Correlations

Image courtesy of Wikimedia Commons


  1. Pandey A, Raza F, Velasco A, et al. Comparison of Morisky Medication Adherence Scale with therapeutic drug monitoring in apparent treatment-resistant hypertension. J Am Soc Hypertension. 2015;9(6):420-426.
  2.  Tamblyn R, Eguale T, Huang A, Winslade N, Doran P. The incidence and determinants of primary nonadherence with prescribed medication in primary care. Ann Intern Med. 2014;160(7):441-450.
  3. Bailey SC, Annis IA, Rueland DS, Locklear AD, Sleath BL, Wolf MS. Development and evaluation of the Measure of Drugs Self-Management. Patient Preference and Adherence. 2015;9:1101-1108.
  4.  Nieuwlaat R, Wilczynski N, Navarro T, et al. Interventions for enhancing medication adherence. Cochrane Database of Systematic Reviews 2014, Issue 11. Art. No.: CD000011. DOI: 10.1002/14651858.CD000011.pub4.

The Blue Phone and the Bow-Tie

March 14, 2014

By Joseph Zakhar

Peer Reviewed

The Patient:

Fate is the sound of a ringing phone.

I, however, am growing to hate the sound.

I’m strangled by the words, by the rough sheets, the silence as a stranger far away connects us, sitting in some room. There’s a tension, an unsettling sense of doom as I count the doctors’ blinks and wait for the “bonjourno.” I hope the translator – the one who lets me and my doctors talk – is somewhere warm, like Texas.

The phone is a harbinger of straight facts, none too good as I sit here with my heart. It’s a necessary coldness. I always sit very, very still as they explain, so precisely, what’s happening. I sit still, afraid to shake the words out of my ears.

Repeat back that I’m agreeing to surgery tomorrow. Do I know what a cabbage procedure is? CABG? AVR? What’s my family’s medical history?


I understand.

A nod can do as much. But only so much. And as an immigrant (and proud Italian), I’ve grown use to the basics: vague gesticulation and mangled speech. But what I’m agreeing to makes me scared for the first time in my many years.

After years in America I continue to wake up at dawn and almost expect to pull on my knee-high rubber boots I wore on the boats back home. My habits have changed now. I’m still waking up in the morning, only to hear the squeak of white Keds on the pristine floors and the gentle rumble of wheelchairs being pushed along, harmonized with the scrapes from the IV poles pushed around by emaciated men in cheap slippers. At night I like to watch the cars swing by the hospital, their headlights all yellow like lazy jaundiced eyes.

I miss the smell of fish from back home. Or maybe it’s that I hate the coldness leeching from the phone into the scent here. It’s like this new place where I can’t talk, can’t communicate, doesn’t want me or my smell here. As if the place is trying its best to ignore me.

Which is why I was so surprised when I heard Italian, not echoing out of the dark recesses of cold plastic, but from a young man. With a bow tie. Not a kid, really. He has a white coat.

So he walked in, the cold blue translator phone ignored in the corner as he sat at the edge of my bed, and just… talked. And in those halting phrases, horrible accents (on his part), wide smiles, and sweeping gestures I establish a new habit.

And as long as I’m here I will talk. To the phone, yes, but also in my rough English. And he’s shown me that he will listen and thinks that I am more than a badly worded turn of phrase. I will see him. I’ve asked him to say hello, or maybe it’s goodbye, before my surgery.

As long as I have breath and can joke and talk about my fish and my women, I will try to tell him. Tell the bow-tie kid.

I will show them all that I am alive.

The Student:

“Adriano and Joe,” he says, laughing and pointing at him, then me. I laugh. I laugh a lot when I spend time with him. And he laughs all the harder to see that a) what he’s said is recognizable and b) it’s funny.

And it means so much to him. He’s had one visitor. One time. In four weeks. No family to speak of. Four weeks of staring out the window. Which means he’s starved for company. And though I relish analyzing his daily lab values and talking to my team, discussing surgery options and catheterization results, there’s something to be said for actively ignoring the heart rate monitor dips, the IV’s swooping with the liquids dancing into his body.

And I live for it. These moments. It’s worth all the times I can almost feel my soul disengage from my body, passing through a bittersweet cloud of coffee stained breath and over my UpToDate printouts on its way out.

I’ve looked over his reports. Pored over his Echocardiogram. His illness isn’t messing around. But there he is. At first, he was all casual Friday. Don’t worry, be happy. A buddy cop movie. And underneath it all was a clawing, desperate sense of isolation. A prisoner of his broken body. A prisoner of his native tongue.

When I first met him, he seemed the embodiment of loneliness. And when I thought of him, he reminded me of the color yellow. A faint statement of a color that claims a small existence, but could be easily swept away by anything else – overwhelmed by other colors.

Or maybe like the letter Q. A broken circle. Rarely used – last pick of the hangman.

And then we talked. And after a week of daily visits, he told me that initially he didn’t want to disappoint me and wasn’t telling me when he was in pain or scared or confused. And then he asked me to talk to him before his procedure. And then asked me to call him when he went home to his one-room apartment. I realized I was no longer a translator or a messenger, by definition never telling my own story. In that brief time I was his confidant, his teacher, his friend.

There’s not much to life but timing. I was there, me with my 3rd generation broken-Italian, when he was there. And thanks to his surgeons, his medical team, and maybe even my friendly chats, he’s alive. In countless ways, in countless colors and letters that disregard language. Alive.

Joseph Zakhar is a 3rd year medical student at NYU School of Medicine

Image courtesy of Wikimedia Commons

Happy New Year!!

December 31, 2013

As we reach the end of another year, we want to wish all our readers a very happy and healthy New Year. And not to be outdone by every other journal, website, magazine, tv show etc, here is our list of the top 10 articles published on Clinical Correlations in 2013, . … Drumroll please…. In no particular order:

1. Reflections on Hurricane Sandy Jessica Taff, MD

2. In Search of a Competitive Advantage: A Primer for the Clinician Treating the Anabolic Steroid User David G. Rosenthal and Robert Gianotti, MD

3. Clinical Misinformation: The Case of Benadryl Causing Dementia Natalie Smith, MD

4. Are We Too Hesitant to Anticoagulate Elderly Patients with Atrial Fibrillation? A Risk-Benefit Analysis Sunny N. Shah, MD

5. The Health Risks and Benefits of Drinking Coffee Anish Parikh, MD

6. Electronic Cigarettes, What We Know So Far Daniel Taupin

7. Decoding the APOL1 Kidney Areeba Sadiq

8. The DLO:DOES FFP Correct INR? Nicole A. Lamparello, MD

9. Does Running Cause Knee Osteoarthritis? Karin Katz, MD

10. Corticosteroids and Prophylaxis:  What Complications Should You Try To Prevent in Patients on Chronic Corticosteroids? Robert Joseph Fakheri, MD

There are so many other outstanding, deserving articles that did not make the list. We encourage you to peruse the archives to get a real sense of all the fascinating internal medicine topics we have published this year. Happy New Year!!



The Diagonal Earlobe Crease: Historical Trivia or a Useful Sign of Coronary Artery Disease?

November 2, 2011

Nicholas Mark, MD & Sarah Buckley, MD

Faculty Peer Reviewed


Publius Aelius Hadrianus, better known as Hadrian, emperor of Rome (117-138 CE), traveler, warrior, and lover of all things Greek, fell ill at the age of 60. He developed progressive edema and episodic epistaxis, fell into a depression soothed by rich food and drink, and succumbed to death within 2 years. The exact cause of Hadrian’s death–whether by heart failure, glomerulonephritis, or even hereditary hemorrhagic telangiectasia–has been a topic of debate among paleopathologists. It was not until 1980 that a crucial clue was found, memorialized in stone busts of the late emperor: he was sculpted with a deep diagonal crease in both earlobes. [1]

Since its first description by Frank in the New England Journal of Medicine in 1973[2], the presence of the diagonal earlobe crease (ELC) has been recognized as a marker of coronary artery disease (CAD). Subsequent studies confirmed the ELC (or Frank’s sign) as a predictor of CAD independent of age, cholesterol, blood pressure, or smoking status. On the other hand, several studies found no correlation between ELC and CAD and suggest that it is simply a marker of advancing age. Over 50 papers have been published regarding this physical diagnosis sign, and for almost 4 decades controversy has raged over its utility. Is the ELC a clinically useful predictor of CAD? In order to answer this question we performed a meta-analysis of all published studies evaluating the role of ELC as a predictor of CAD.


Published articles, abstracts, and letters were obtained using the search term “Ear Lobe Crease.” Raw data regarding prevalence of ELC and CAD were collected and analyzed to calculate sensitivity, specificity, and likelihood ratios (LR). Significance was determined using the Fisher exact test and Chi squared test (depending on the number of patients in the study), and P values and 95% confidence intervals (CI) were calculated.


There was significant variation in the design of the studies, ranging from large population screening studies to smaller studies looking at patients undergoing angiography for suspected CAD. The majority of studies found that ELC was a statistically significant predictor of CAD: the results of 6 of the 22 studies analyzed were not significant, and the remaining 16 studies demonstrated varying degrees of predictive value, with likelihood ratios ranging from of 1.33 to 9.20 (see figure 1). Upon pooling all data, we found an overall sensitivity of 60.4% and specificity of 74.4%. We found that the presence of ELC has a LR of 2.37 (CI of 2.26 to 2.48) for predicting CAD. We analyzed the subset of studies that included only cardiac patients and found that the utility of ELC was lower in this group than in an unselected patient population (LR of 1.88 vs 2.44). Similarly, when the higher-risk diabetic population of the Fremantle Diabetes study was excluded, the LR was slightly higher.


Frank suggested that the presence of “a prominent crease in a lobule portion of the auricle” may be indicative of small vessel pathology, possibly explaining the correlation with CAD.[1] Shoenfield and colleagues performed histological examination of earlobe creases and found substantial thickening of the arteriolar walls relative to controls without ELC.[25]

Many authors performed multivariate analysis to ascertain if ELC is truly an independent marker of CAD risk or just a surrogate for other known risk factors. Several studies found that though ELC correlates with CAD, it is independent of other CAD risk factors (hypertension, hyperlipidemia, diabetes, and smoking).[13,15].

Other studies reported that that the LR of bilateral ELC is higher than that of unilateral ELC [15], and that depth of the ELC portends a greater likelihood of CAD. It has also been proposed that the presence of additional factors, such as earlobe hair [17], could further increase the utility of the sign.

The utility of ELC may vary depending on several factors. Age appears to be a significant potential confounder, and indeed the incidence of ELC increases with age in every study. However, as expected, the incidence of CAD also increases with age, and it is unclear at what ages the utility of the sign is highest. Several studies performed subgroup analyses looking at the utility of the ELC in patients of different ages with conflicting results. In general, though, the studies found that the presence of ELC has some predictive value across all age ranges.

A few studies have reported negative results in the context of specific ethnic groups. Fisher and colleagues found no significant relationship between ELC and CAD in American Indians [11], and Rhoads and colleagues found no relationship in Japanese-Americans living in Hawaii.[8] Overall, the published incidence of ELC seems to vary significantly among different populations; the significance of this is not clear.

The utility of ELC may also be lower in patients who have a higher pretest probability of CAD, such as in the studies of patients undergoing angiograms for assessment of suspected CAD.[3,17,22] Furthermore, the Fremantle diabetes study suggested that among patients with one of the most important CAD risk factors, diabetes, there is no correlation between ELC and CAD.[21]

Given these limitations, what is the utility of ELC relative to the more established CAD risk factors such as diabetes, hypertension, hyperlipidemia, and smoking? Unlike the modifiable risk factors, which can be addressed by medical management, ELC is only a marker for coronary disease. A recent epidemiological study by Greenland and colleagues [26] and review by Weissler in JAMA [27] showed that the predictive value of these traditional risk factors for predicting CAD complications (MI or death) was quite low (LR ranging from 1.07 to 1.39). Though our meta-analysis looked at ELC as a predictor of CAD (rather than complications of CAD), we find that while the LR of ELC for predicting CAD is modest (LR of 2.37, CI of 2.13 – 2.88), it is significantly higher than that of traditional risk factors. Although we find that the presence of ELC is neither especially sensitive nor specific for CAD, when compared to other known risk factors it does appear to be useful. Thus, we propose that ELC may be a useful additional marker for identifying patients with CAD.

Perhaps if Hadrian were alive today, his sculptors would not have been the only ones to take note of his ears.


Drs. Nicholas Mark and Sarah Buckley are former students of NYU School of Medicine

Reviewed by Beno Oppenheimer, MD, Assistant Professor Medicine, Division Pulmonary/Critical Care, Course director Introduction to Bedside Diagnosis, NYU School of Medicine

Image courtesy of Wikimedia Commons (Hadrian, emperor of Rome)


1. Petrakis NL. Diagonal earlobe creases, type A behavior and the death of Emperor Hadrian. West J Med. 1980;132(1):87–91.

2. Frank ST. Aural sign of coronary-artery disease. N Engl J Med. 1973;289(6):327-328.

3. Lichstein E, Chadda KD, Naik D, Gupta PK. Diagonal ear-lobe crease: prevalence and implications as a coronary risk factor. N Engl J Med. 1974;290(11):615-616.

4. Mehta J, Hamby RI. Letter: Diagonal ear-lobe crease as a coronary risk factor. N Engl J Med. 1974. 291(5):260.

5. Christiansen JS, Mathiesen B, Andersen AR, Calberg H. Letter: Diagonal ear-lobe crease in coronary heart disease. N Engl J Med. 1975;293(6):308-309.

6. Sprague DH. Diagonal ear-lobe crease as an indicator of operative risk. Anesthesiology. 1976;45(3):362-364.

7. Doering C, Ruhsenberger C, Phillips DS. Ear lobe creases and heart disease. J Am Geriatr Soc. 1977;25(4):183-185.

8. Rhoads GG, Yano K. Ear-lobe crease and coronary-artery heart disease. Ann Intern Med. 1977;87(2):245.

9. Kaukola S. The diagonal ear-lobe crease, a physical sign associated with coronary heart disease. Acta Med Scand Suppl. 1978;619:1-49.

10. Wermut W, Jaszczenko S, Ruszel A. Ear lobe crease as a risk factor in coronary disease. Wiad Lek. 1980;33(6):435-438.

11. Fisher JR. Sievers ML. Ear-lobe crease in American Indians. Ann Intern Med. 1980;93(3):512.

12. Kaukola S. The diagonal ear-lobe crease, heredity and coronary heart disease. Acta Med Scand Suppl. 1982;668:60-63.

13. Elliott WJ. Ear lobe crease and coronary artery disease. 1,000 patients and review of the literature. Am J Med. 1983;75(6):1024-1032.

14. Wagner RF Jr, Reinfeld HB, Wagner KD, et al. Ear-canal hair and the ear-lobe crease as predictors for coronary-artery disease. N Engl J Med. 1984;311(20):1317-1318.

15. Gu?iu I, el Rifai C, Mallozi M. Relation between diagonal ear lobe crease and ischemic chronic heart disease and the factors of coronary risk. Med Interne. 1986;24(2):111-116.

16. Gibson TC, Ashikaga T. The ear lobe crease sign and coronary artery disease in aortic stenosis. Clin Cardiol. 1986;9(8):388-390.

17. Verma SK, Khamesra R, Bordia A. Ear-lobe crease and ear-canal hair as predictors of coronary artery disease in Indian population. Indian J Chest Dis Allied Sci. 1988;30(3):189-196.

18. Kenny DJ, Gilligan D. Ear lobe crease and coronary artery disease in patients undergoing coronary arteriography. Cardiology. 1989;76(4):293-298.

19. Miri? D, Rumboldt Z, Pavi? M, Kuzmani? A, Bagatin J. The role of diagonal ear lobe crease in the clinical evaluation of coronary risk. Lijec Vjesn. 1990;112(7-8):206-207.

20. Moraes D, McCormack P, Tyrrell J, Feely J. Ear lobe crease and coronary heart disease. Ir Med J. 1992;85(4):131-132.

21. Motamed M, Pelekoudas N. The predictive value of diagonal ear-lobe crease sign. Int J Clin Pract. 1998;52(5):305-306.

22. Davis TM, Balme M, Jackson D, Stuccio G, Bruce DG. The diagonal ear lobe crease (Frank’s sign) is not associated with coronary artery disease or retinopathy in type 2 diabetes: the Fremantle Diabetes Study. Aust N Z J Med. 2000;30(5):573-577.

23. Dytfeld D, Le?na J, Protasewicz A, Sarnowski W, Dyszkiewicz W, Paradowski S. Ear lobe crease as a factor of potential risk for coronary artery disease?–World news review and own research. Pol Arch Med Wewn. 2002;108(1):633-638.

24. Bahcelioglu M, Isik AF, Demirel B, Senol E, Aycan S. The diagonal ear-lobe crease. As sign of some diseases. Saudi Med J. 2005;26(6):947-951.

25. Shoenfeld Y, Mor R, Weinberger A, Avidor I, Pinkhas J. Diagonal ear lobe crease and coronary risk factors. J Am Geriatr Soc. 1980;28(4):184-187.

26. Greenland P, Knoll MD, Stamler J, et al. Major risk factors as antecedents of fatal and nonfatal coronary heart disease events. JAMA. 2003;290(7):891-897.

27. Weissler AM. Traditional risk factors for coronary heart disease. JAMA. 2004;291(3):299-300.

The Stories My Senses Tell Me

May 20, 2011

By David Ellenberg

Faculty Peer Reviewed

Mr. A is a 91 year old male with a history of hypertension and two myocardial infarctions. He presented with shortness of breath and worsening lower extremity swelling and was subsequently admitted to the medicine floor for a CHF exacerbation. While on the floor, he also complained of painful pressure ulcers on his lower buttocks. He was diuresed and given oxygen by nasal cannula during his 48-hour hospital stay, and his ulcers were cleaned and monitored for infection. He was discharged with a new pharmacologic regimen.

On the second morning of his hospital stay, I asked Mr. A how his night was at the precise moment when my inexperienced hand placed a stethoscope on his chest. My nervous attempt to weave non-scripted questions into the patient encounter resulted in indistinct heart sounds laced with the vibrations from his voice. I had tried too quickly to learn as much about him as I could and, in the process, heard neither what he, nor his heart, had to tell me. I was taken aback, fearing that my oversight had crudely displayed my uncertainty. More significantly, I had a sudden awareness of the wide variety of tools we use to get to know people as I tried to tease out the uniqueness of his emotional and physical rhythms. I appreciated how perceptive all of our senses must become, and how sensitive they must remain, when assessing patients.

The most challenging, and yet fascinating, aspect of transitioning between knowledge and practice is isolating and synthesizing the different ways we examine people. I had felt Mr. A’s edematous ankles, seen subtle ECG changes, smelled his infected ulcers, and heard his extra heart sounds, pulmonary vibrations, and most importantly, his words. Each sensory element of my daily inspection formed my version of his clinical and personal narrative. I had been weighing all my senses, relating them to each other, and looking for combinations and sequences. My senses and the spoken—and unspoken—words of my patient coalesced to form the stories of his illness and hospital course. What we sense from our own experiences with our patients adds immeasurably to the stories we tell, retell, and condense on rounds and for sign-outs. These stories begin as simple accounts of our patients’ activities, but over time, we make them our own.

The strict verbal relay of information between and within teams can only bear so much, though. We prize accuracy and objectivity, but these realities are shaped by the experiences with those under our care. I remember examining a different patient last month and appreciating a subtle, questionable murmur which only half the team had heard. When I wrote my daily progress note later that day, my intern, who had not heard the murmur, made sure I included the abnormal sound as a finding of my physical exam. “It was what you heard,” he explained to me. In a field full of inconsistencies and wonderful surprises, where a possibility can regularly overshadow a definite, the notions of formal truths and certainties can fall by the wayside. There are no absolutes in medicine. It is only through these personalized and individual senses that we can connect to our patients on a more fundamental level and that their stories become our own.

I love that about medicine. I can read stacks of journals and participate in conferences and become genuinely absorbed by Mr. A’s clinical and pathological details. Yet, only my unique experiences with him enable me to add to or appreciate them. During the two days I spent with Mr. A in the hospital, I touched, saw, smelled, and listened. I even tasted his Ensure one morning when he was expressing to me what the hospital felt like to him. I heard his stories and experienced his illness through words, silences, and my own memories of my grandparents as they were stricken with heart disease. My story of Mr. A is not complete, or in my control, without the contribution of each sense. We add authenticity to the stories we tell when they are infused with our own senses and perceptions of their characters. Mr. A’s story remains his own, and was not compromised by my version that I recounted each morning.

My experience with him was personal, and it allowed me to be able to see the patient as a character in the stories he told. I came to understand him based on our interpersonal encounters, and not on what I read in a chart. But the value transcended merely understanding Mr. A as an individual. I came to respect each sense’s ability to contribute to the story and appreciate that the very means with which I navigate the world and my surroundings can connect me to another human being. The story of Mr. A, which I reproduced several times per day, grew out of that bond. Senses, though, require their own time and space to allow for that. I initially tried to sense so many tones at once that I heard none of them. When I finally removed the stethoscope from his chest, Mr. A would tell me his own stories of the sights, sounds, and smells of the army base kitchen where he once spent so much time. And I felt much more connected, much more privileged,  much more involved when his own senses were part of his storytelling.

David Ellenberg is a 4th year medical student at NYU Langone Medical Center

Peer reviewed by Ishmeal Bradley, section editor, Clinical Correlations

Image courtesy of Wikimedia Commons


Tsunamigenesis and Korotkoff’s Sounds

January 26, 2011

By Irene Isabel Payad Lim, MD and Michael Ford, MD

Faculty Peer Reviewed

The February 27, 2010 earthquake in Chile measured 8.8 on the Richter scale and displaced nearly 2 million people[1].  It also threatened to generate tsunamis that were predicted to hit the coasts of Japan, the Philippines, and Hawaii some 7000 miles away.  In this instance, the waves set off by the Chilean earthquake dissipated relatively harmlessly.  On the day after Christmas 2004, however, 14 countries bordering the Indian Ocean were not as lucky. Thailand, Sri Lanka, and India were the hardest hit[2].  The tsunami decimated their shores without warning and killed 230,000 people.  Originating from an offshore earthquake near the Indonesian island of Sumatra, the tsunami ended its destructive path on the East African shore, some 3000 miles away and seven hours after it began (Figure 1)[3].


Figure 1. Tourists fleeing the tsunami as it comes ashore at Hat Rai Lay Beach, Thailand.             Agence France-Presse

In the aftermath of the Chilean quake, tsunamis were a prominent topic of conversation on our morning work rounds.  Tsunamis and their formation may seem like an esoteric subject for daily rounds on the Bellevue medicine ward, and even more so for an internal medicine blog.  The qualities of these powerful waves, however, have a connection with a daily part of the practice of medicine, blood pressure measurement.  Once an art form touted by prominent scientists such as Ludwig and Korotkoff, blood pressure measurement has become such a commonplace practice that is often taken for granted.  Upon cursory review, tsunamis may appear far removed from our everyday lives, but the turbulent tides of the Earth’s oceans share many similarities to the metaphorical tides brewing within each of us.

Tsunamis form in four stages: initiation, split, amplification and runup.  The initiation begins with a submarine earthquake, during which the seafloor oscillates up and down[4]due to the increased pressure emitting from below.  These vertical movements of the Earth’s crust can be quite large and cause the sea water to be displaced above the mean sea level (Figure 2).  These events are so intimately tied that the main determinant of the initial size of a tsunami is the amount of vertical sea floor deformation[4]. The level at which the water is pushed above mean sea level is translated into the water’s potential energy.  This newly-acquired potential energy is then transferred to kinetic energy, which is directly related to the water velocity squared.  The following equations govern this phenomenon5:


KE = ( ½ ) (m) (v2)


where Δ = “change in”, PE = “potential energy”, KE = “kinetic energy”, m = “mass”, and v = “velocity”.

Figure 2: Seafloor movements in submarine earthquake result in upward seawater displacement [6]


As the tsunami progresses, the water gains its velocity not just from the kinetic energy gained from the earthquake but also from conditions surrounding its formation.  Specifically, the tsunami’s speed varies directly with the square root of the water depth, thus making deep ocean tsunamis faster.  In the Pacific Ocean, where certain points can reach a depth of 4000 meters, tsunamis can travel at speeds upward of 700 km/h, with wavelengths of more than 100 km.  These parameters make a deep ocean tsunami akin to a very long, shelf of water moving across the ocean at the speed of a jet airplane!  By Reynolds equation, the speed at which the water travels determines whether fluid flow is laminar (smooth) or turbulent (chaotic) [5].  The greater the velocity, the greater the Reynolds number, thus resulting in turbulent flow.  The Reynolds equation is defined belo[5]:



Sea water can be imagined as fluid within an imaginary pipe in which the atmosphere and seafloor are the boundaries above and below, respectively.  In a pipe, a Reynolds number above about 4,000 corresponds to turbulent flow, while a Reynolds number below 2,100 indicates laminar flow [5].  The region in between (2,100 < Re < 4,000) is called the transition region. In turbulent flow, unsteady fluid vortices appear and interact with each other[5].  (While ocean water can be conceived of as flowing in a pipe, the depth of the ocean floor varies considerably across its length, making it difficult to calculate pipe hydraulic diameter; thus, a determination of Reynold’s number for ocean waves is beyond the scope of this essay.)

The same principles that govern tsunami formation can be applied to the flow states of our own blood.  Blood pressure, an indicator of the circulatory flow in the human body, can be measured by auscultation, automated (oscilloscopic) and invasive techniques.  Though automated oscilloscopes are increasingly used in both the clinic and the hospital, the auscultatory method with mercury sphygmomanometers provides the most accurate non-invasive measurement of blood pressure[7].

In 1905, Nikolai Korotkoff was the first to observe the production of unique sounds by arterial constriction[8]. The auscultatory method relies on the first and fifth Korotkoff’s sounds, which are dependent on flow states that change at specific times[9].  The appearance of the first, faint tapping sounds is clinically correlated with the systolic blood pressure while the point at which all sounds disappear correlates with the diastolic blood pressure[9].  A recent study by Sykes, et al. used ultrasonics to definitively determine the cause of each sound[10].  They found that the first Korotkoff sound, known as Phase 1 or K1, coincides with large amplitude movements of the brachial artery as it begins to open from its collapsed position under external pressure from the cuff.  This is analogous to the external pressure exerted by the submarine earthquake onto the body of water above it.

As the cuff is deflated, flow is restored through the brachial artery.  The evolution of this flow is divided into further stages, Phase 2 through 5 (K2 to K5).  K2 was found to coincide with high velocity jets of blood flowing through the brachial artery[10].  One could imagine that the blood, like sea water during tsunami formation, gains potential energy (and thus kinetic energy) from its displacement secondary to brachial artery narrowing due to increased pressure from the cuff.  The kinetic energy then translates into an increase in velocity.  Indeed, there is high velocity flow at this stage[9] and again, the Reynolds number can approach a value predicting turbulent flow.  According to Sykes, et al, the Reynolds number correctly predicts turbulent flow as flow during K2 contains vortices[10], a characteristic of turbulence in fluids.  On auscultation, K2 correlates with the onset of the murmur phase[10].  Clinically, any murmurs have been associated with turbulent flow (i.e. bruits in carotid artery stenosis), which is supported by this evidence.

Conversely, the absence of turbulent flow (and thus, presence of laminar flow) has been clinically associated with the absence of murmurs.  In relation to Korotkoff’s sounds, the disappearance of sounds (at K5) has been historically related to the diastolic pressure.  The ultrasonic study showed that at K5 there was a decrease in amplitude of arterial wall movements and the return of laminar flow10.  The restoration of laminar flow, in turn, implies a decrease in velocity (by calculation of Reynolds number).  These results suggest that disturbance of flow, as defined in stages by Korotkoff’s sounds, are directly related to brachial artery wall disturbances secondary to blood pressure cuff inflation and deflation.  Similarly, disturbance of flow that results in tsunamis are directly related to seafloor disturbances secondary to submarine earthquakes.

Alternatively, one can imagine that sea water within an imaginary pipe (again, with the atmosphere and the sea floor as its superior and inferior boundaries), experiences the creation of a pressure gradient when the sea floor oscillates during an underwater earthquake.  In this manner, the physics behind tsunami formation and blood pressure measurement can also be proven with Darcy’s law of blood flow.  This law states that blood circulation is dependent on a pressure gradient [5].  The mathematical equation describing this relationship is defined below:

F = DP/R

where F = blood flow, DP = pressure gradient, R = resistance

During blood pressure measurement, the inflated pressure cuff rapidly introduces a pressure gradient between two points along the brachial artery while simultaneously decreasing the artery’s radius (Figure 3), resulting in the turbulent flow of blood, which is detected by Korotkoff’s sounds.  Similarly, a submarine earthquake introduces a pressure gradient through the elevation of water pressure above the mean sea level.  The flow then changes from laminar to turbulent, and depending on other inciting factors such as water depth, can result in tsunamis.  The only difference in the two situations is the causative, inciting factor for the presence of a pressure gradient.



Figure 3: Velocity in a blood vessel.  Areas of high velocity and turbulent flow occur at points of high pressure and decreased radius [9].

Though tsunamis have enormous potential for destruction, the disturbance of blood flow in blood pressure measurement thankfully does not.  This contrast can be attributed to the stark difference in magnitude of energy that is conferred, even though the production of that energy is notably similar in the two situations, as shown above.  Ultimately, by employing the fundamental principles of fluid mechanics, the brachial artery wall disturbances in blood pressure measurement can be likened to disturbances of the sea floor in tsunami formation, leading to disturbances in blood and sea water flow, respectively.  To our knowledge, we are the first to propose this relationship between a great natural phenomenon (tsunamigenesis) and the mechanics behind a key vital sign (blood pressure measurement).

Commentary: An Alternative View

By Leon Axel M.D., Ph.D. Professor; Departments of Radiology, Medicine, Physiology and Neuroscience, NYU Langone Medical Center

This is an entertainingly written essay, with a striking photo of an incoming tsunami, and it contains a nice review of some of the physical aspects of fluid flow in both tsunamis and Korotkoff’s sounds of blood pressure measurements. However, I am skeptical of some of the comparisons that are made between the rather different phenomena of tsunamis and the Korotkoff’s sounds.

It is certainly true that they both involve moving fluids, and they thus must be subject to the same underlying physical properties of fluid flow. However, the scales of the energies involved are greatly different in the two phenomena, and the relative roles of turbulence are also quite different.

As I understand it, in the generation of tsunamis, the unique aspect of them that differentiates them from an ordinary wave on the ocean, is that they involve a local vertical motion of essentially the whole depth of the ocean, a result of the immense forces produced by the lifting of the sea floor. In contrast, the motion of an ordinary ocean wave just involves the surface of the ocean. Thus, the kinetic energy involved in a tsunami is a result of not only the high speed at which it travels, but the fact that it is the whole depth of the ocean that is moving. In fact, the wave is actually apparently quite small in amplitude at the surface, while it is still in the open ocean. The Reynolds number is not really relevant here, as the fluid motion is not very turbulent when it is travelling across the open ocean. (In any case, the only boundary that would be relevant for drag on the fluid in calculating the Reynolds number would be the ocean floor, as the motion at the surface is unconstrained). As the wave approaches the shallower depths near the shore, all of the kinetic energy stored in the deep wave motion shows up as a growing amplitude of the wave at the surface, with the resulting great height of the wave at the water’s edge, with its associated destructive power. The breaking wave is turbulent, of course, but so is an ordinary breaking wave; it is not the turbulence, per se, that distinguishes a tsunami, but rather the amount of energy involved in it.

In the Korotkoff’s sounds, the source of the audible sounds is the turbulence that is associated with flow through a narrowed artery in certain flow regimes, as nicely described in the essay. However, source of the energy dissipated in that turbulence is the contracting heart upstream, not the compressing cuff at the local constriction of the artery. Thus, the analogy being made between the forces of the blood pressure cuff and the lifting ocean floor as causes of the two phenomena is, I feel, confusing.

In summary, the descriptions of the two separate phenomena are reasonably well done in the essay, but I feel that the comparison that is being made between them is a bit far fetched and potentially confusing.

Dr. Lim is a 4th year medical student at NYU School of Medicine

Dr. Ford is a 3rd year resident at NYU Langone Medical Center

Peer review and commentary by Leon Axel M.D., Ph.D. Professor; Departments of Radiology, Medicine, Physiology and Neuroscience, NYU Langone Medical Center

Image courtesy of Wikimedia Commons (2004 Indonesia Tsunami)


1. Barrionuevo A. and Robbins L. 1.5 Million Displaced After Chile Quake. New York Times (New York Ed.). 2010 Feb 28: Sect. A:1.

2. Cumming-Bruce N. and Robertson C. Most Powerful Quake in 40 Years Triggers Death and Destruction. New York Times (International Ed.). 2004 Dec 26.

3. Revkin AC. How Scientists and Victims Watched Helplessly. New York Times (International Ed.). 2004 Dec 31.

4.  Levin B, Nosov M. Physics of Tsunamis. Netherlands: Springer Science, 2009: 1-10, 77-94.

5. Fox RW, McDonald AT and Pritchard PJ. ­Introduction to Fluid Mechanics. 7th ed. New York, NY: Wiley, 2008: 35-40, 328-336.

6.  Valdes R and Halabrin N.  “How Tsunamis Work.”  04 January 2005.>  7 March 2010.

7. Canzanello VJ, Jensen PL and Schwartz GL. Are aneroid sphygmomanometers accurate in hospital and clinic settings?. Arch Intern Med. 2001 Mar 12;161(5):729-731.

8. Pickering TG, Hall JE, Appel LJ, et al. Recommendations for Blood Pressure Measurement in Humans and Experimental Animals: Part 1: Blood Pressure Measurement in Humans : a statement for professionals from the Subcommittee of Professional and Public Education of the American Heart Association Council on High Blood Pressure Research.. Circulation. 2005 Feb 8;111(5):697-716.

9.  Barrett KE, Barman SM, Boitano S, Brooks H, “Chapter 32. Blood as a Circulatory Fluid & the Dynamics of Blood & Lymph Flow” (Chapter). Ganong’s Review of Medical Physiology. 23rd ed. New York, NY: McGraw-Hill Medical, 2009: 521-554.

10. Sykes DA, McCarty K, Mulkerrin E, et al. Correlation between Korotkoff’s sounds and ultrasonics of the brachial artery in healthy and normotensive subjects. Clin Phys Physiol Meas. 1991 Nov 1;12(4):327-331.

Bedside to Bench: Clubbing Revisited

April 3, 2009

Commentary by Judith Brenner MD, Associate Editor, Clinical Correlations 

Faculty Peer Reviewed 

For an internist, discovering a patient with clubbing is so rewarding since it appeals to the core of our profession, a profession which can often be very similar to that of a detective. The physical finding of clubbing was first described by the ancient Greeks, who recognized it to be a clue to much more.

When a clinician discovers clubbing of the fingers, he must consider that hypoxemia may be present, whether secondary to a cardiac or pulmonary process. In fact, more than 90% of clubbing is pathologic, with the remainder being a benign familial condition.

How do we recognize clubbing on physical exam?

Let’s begin at the bedside. One can judge finger clubbing in many ways. The simplest way is to think of the clubbed finger is as a “drumstick digit”. There is an increase in the curvature of the nail and a general rounding at the tip. However, curvature is difficult to measure at the bedside. Most physicians look for the “Shamroth Sign”. This sign, named for the doctor who first described it in 1976, is performed by looking for loss of the diamond that is usually formed in non-clubbed fingers when the dorsal surfaces of the distal phalanx of the right and left fingers are apposed. Unfortunately, though commonly considered to be a “standard” in terms of diagnosis, this sign has never been rigorously studied.

Another way to determine if clubbing is present involves use of the “phalangeal depth ratio.” This ratio compares the “distal phalangeal depth” with the “interphalangeal depth”. In normal individuals, the DPD:IPD ratio is <1. However, in clubbed fingers, the distal portion is thicker and thus, the ratio of DPD:IPD is >1.

The last finding to consider is the hyponychial angle. The normal angle is 180 degrees. With clubbing the curvature of the nail increases and the angle thus increases. An angle greater than 190 degrees is considered consistent with clubbing.

Unfortunately determining the evidence for these findings is difficult.  No sensitivities, specificities or likelihood ratios are available.  Why? Simply because there is no gold standard.   Here’s where experience counts.  A clinician can look at a finger, observe that its appearance is not “normal” and then start to ask the appropriate questions.

Is clubbing really a marker of hypoxic disease?

The answer is: yes.  In one study of 350 patients with clubbed fingers, 80% had underlying respiratory disorders, including tumors, abscesses, cystic fibrosis, and interstitial fibrosis.  An additional 10-15% had cyanotic heart disease, endocarditis, and even thyroid disease and inflammatory bowel disease.  Only 5% were found to be either idiopathic or hereditary.

What is the unifying pathology in all of these disorders?

On a molecular level, the pathophysiology has been studied by investigating rare individuals with familial hypertrophic osteoarthropathy.  In these patients, mutations of HPGD, a prostaglandin E2 catabolizing enzyme, has been identified.  This mutation results in elevated prostaglandin levels.  Although the familial form is rare, the secondary forms of hypertrophic osteoarthropathy (HO) are more common and clubbing is often the first clue.  The clinical commonality in many patients with secondary HO is right to left shunting.  Ordinarily, prostaglandin E2 (PGE2) is metabolized in the lung.  The hypotheseis is that, with shunting, the proper metabolism is prevented.  As a result, PGE2 levels are elevated and elevated prostaglandin levels result in platelet activation. 

Activated platelets are returned to the systemic circulation and are thought to lodge in the distal phalanges, releasing their growth factors.  While trapped, the platelets release growth factors, which leads to fibrovascular proliferation, ultimately manifesting in what we clinically call clubbing.

Though compelling, these explanations are all still just hypotheses.  While many diseases associated with clubbing have shunting in common, several do not, such as Graves’ Disease and inflammatory bowel disease.  The pathophysiology in these cases remains unclear.

The Bottom line:

Clubbing is indeed a clinical clue to an underlying disorder and thus must be taken seriously when recognized.  It is diagnosed at the bedside where simple observations are made.  When present, a search for diseases of the lungs or heart that cause right to left shunting is warranted.  If cardiac and pulmonary etiologies are ruled out, one can consider other diseases that are marked by platelet excess, such as inflammatory bowel disease.  Inheritance (<5%), since it is so rare, should be accepted as the etiology only as a diagnosis of exclusion.

Coggins, KG et al. The hippocratic finger points the blame at PGE.   Nature Genetics 2008;40:691-2.

Meyers, KA et al. Does this Patient Have Clubbing? JAMA 2001;286:341-347.

Uppal, S, et al.  Mutations in 15-hydroxyprostaglandin dehydrogenase cause
Primary Hypertrophic Osteoarthropathy.  Nature Genetics 2008;40:789-93.

McGee, S.  Evidence Based Physical Diagnosis (2nd edition).

Reviewed by Nishay Chitkara MD, NYU Division of Pulmonary and Critical Care Medicine

Grand Rounds: “Evidence-Based Physical Exam”

December 10, 2008

Commentary by Steven Giovannone MD, PGY-2

Please also see the clinical vignette presented before last week’s grand rounds.

This week’s Medical Grand Rounds was given by Dr. Steven McGee from the University of Washington School of Medicine, author of the well-known textbook “Evidence-Based Physical Diagnosis”.  His lecture, entitled “Evidence-Based Physical Exam,” targeted the utility of the physical exam in an era of increasingly used alternative diagnostic modalities, especially imaging procedures.

In his talk, Dr. McGee first noted that for several medical conditions, physical diagnosis is still the gold standard despite advances in imaging modalities, citing examples such as cellulitis, diabetic retinopathy, and amyotrophic lateral sclerosis.  Alternatively, many diagnoses are now made on grounds other than the exam, for example chest radiography in the diagnosis of pneumonia.  With new “gold standards” for making a diagnosis we are now able to study how effective traditional physical exam maneuvers are in predicting the presence or absence of a medical condition.  The quantification of the accuracy of such physical exam maneuvers is the basis of his textbook.

Dr. McGee argues the most useful measure of diagnostic accuracy is the likelihood ratio (LR), defined as the probability of a physical finding in a patient with a condition divided by the probability of that same finding in a patient without that condition.  For any physical finding, a high likelihood ratio indicates that finding increases the probability of disease and a low likelihood ratio decreases it.  Therefore, tests with likelihood ratios close to 1 can be interpreted as clinically useless because they do not clinically affect the pre-test probability of disease.

In his initial example of ascites using abdominal ultrasound as the gold standard, studies have shown that a positive fluid wave and presence of edema have high likelihood ratios for the presence of ascites (LR of 5 and LR of 3.8 respectively).  On the other hand, other traditional measures such as presence of bulging flanks and shifting dullness were less helpful.

Four more scenarios were presented.  The first asked what elements of the physical exam are helpful in determining the presence of elevated left atrial (LA) pressures in patients presenting with dyspnea.  Using cardiac catheterization as the gold standard, findings such as an S3 (LR 5.7), displaced apical impulse (LR 5.8), and a positive abdominojugular test (LR 8.0) were highly predictive of elevated LA pressures.  Other findings such as edema and crackles had likelihood ratios closer to 1.  The proposed explanation for why crackles are less helpful is simply because one would expect to hear crackles in other conditions that do not elevate LA pressure, including interstitial fibrosis, pneumonia, and COPD.

The other scenarios presented were peritonitis, rotator cuff tear, and pneumonia.  Again, the evidence behind several physical exam findings was presented.  In the example of pneumonia (using a chest x-ray as the gold standard), many findings were suggestive of the presence of pneumonia, including cachexia (LR 4.0), bronchial breath sounds (LR 3.0), and egophony (LR 4.1).  However, in this example no single finding was helpful in excluding pneumonia (LRs all near 1).  In order to more effectively exclude pneumonia, it is possible to combine multiple findings.  The Heckerling score is a constellation of 5 findings (temperature greater than 37.8, tachycardia, crackles, decreased breath sounds, absence of asthma).  If there are 0 or 1 findings present, the likelihood ratio of pneumonia is 0.3.

In summary, Dr. McGee reiterated there are several ways to approach the bedside exam when the gold standard for a condition is an alternative modality.  He proposes that instead of abandoning the physical exam, we can use likelihood ratios to quantify the accuracy of our approach.  By using exam maneuvers that have evidence for their ability to make a diagnosis, we can significantly affect our pre-test suspicion of disease.  To be knowledgeable of such an approach will allow us as physicians to be more confident in our findings, and will ultimately boil down to better patient care.

The Rational Clinical Examination: Does This Patient with Diabetes Have Osteomyelitis of the Lower Extremity?

March 28, 2008

Commentary by Judith Brenner MD, Associate Program Director, NYU Internal Medicine Residency Program

The most recent installment in JAMA’s Rational Clinical Exam Series seeks to determine the accuracy of the history, physical exam, radiology and laboratory in making the diagnosis of osteomyelitis in diabetics. This is relevant given its frequency of occurrence and its cost and since the gold standard for diagnosis, namely a bone biopsy and culture, is less than optimal for a variety of reasons.

ulcer.jpgLess than 10% of the nearly 300 articles reviewed were included for review by the authors and of the 10%, most were graded in the mid categories for quality. Nonetheless, some important findings were reported (highlighted in green).

Finding Likelihood Ratio Negative Likelihood Ratio
Ulcers >2 cm2 7.2 (CI 1.1-49) 0.48 (CI 0.31-.076)
Positive “probe to bone” test* 6.4 (CI 3.6-11) 0.39 (CI 0.20-0.76)
ESR >70 11 (CI 1.6-79) 0.34 (CI 0.06-1.9)
Abnormal plain X-ray** 2.3 (CI 1.6-3.3) 0.63 (CI 0.51-0.78)
Abnormal MRI 3.8 (CI 2.5-5.8) 0.14 (CI 0.08-0.26)

*Probe to bone test: the examiner gently and in a sterile fashion, probes the ulcer with a steel probe to determine if the probe can advance to bone

**abnormal X-Ray findings include: focal loss of trabecular pattern, periosteal reaction, and frank bone destruction. 2 or 3 views can be selected

For example, the interpretation is as follows:

The LR (sensitivity/1-specificity) tells you how much the odds of a disease will increase based on a positive result.

So, if you determine the pre-test probability to be 15% and the patient has an ulcer >2.0cm2 then the post-test probability would rise to 56%.


Pre-test odds = pre-test probability /(1-pre-test probability)= 0.15/0.85=0.18 Post-test odds=pre-test odds * LR = 0.18 * 7.2 = 1.3
Post-test probability = post-test odds / (post test odds+1) = 1.3/2.3= 0.56 = 56%

or, more simply, use a Likelihood ratio nomogram Read more »

Bedside Rounds Series: Goiter

February 21, 2008

Commentary by Kristin Remus MD, PGY-3, and Judith Brenner MD, Associate Program Director, NYU Internal Medicine Residency Program

The purpose of the physical diagnosis series is to review the origins, pathophysiology and actual techniques of common physical diagnoses. Our hopes are to renew interest in an area of medicine that is occasionally overlooked due to the current use of laboratory values and imaging tests frequently used to make a diagnosis.

In the industrialized world, enlarged thyroid, or “goiter”, occurs in 10% of women and 2% of men. The most common causes of goiter in the industrialized world are Hashimoto’s thyroiditis and Graves’ disease. However, most patients with goiter are euthyroid. Goiters are classified as either nontoxic (euthyroid) or toxic (hyper or hypothyroid functional state), as well as simple (one enlarged nodule) or multinodular (many enlarged nodules). For example, a simple nontoxic goiter is an asymptomatic enlargement of the thyroid (i.e., the metabolic state is normal). In contrast, a toxic uninodular or multinodular goiter is an autonomously functioning goiter (i.e. the metabolic state is abnormal). A goiter is an important finding, and small goiters are frequently missed on examination of the thyroid.

Physical Exam of the Normal Thyroid
Physical exam of the thyroid employs the basic principles of identification and inspection, followed by palpation. To identify the thyroid gland, find the laryngeal prominence of the thyroid cartilage which is the most conspicuous prominence in the neck. Next, locate the cricoid cartilage. The isthmus of the thyroid lies just below the cricoid cartilage. Note that these structures create a “V” shape in the neck, making for easy identification. Next ask the patient to extend their neck. This will lift the trachea and stretch the skin against the thyroid, allowing for better visualization.

Then, inspect the patient’s neck from the side. You should see a straight line from the thyroid cartilage to the cricoid cartilage. If there is an anterior bowing of this line, this suggests a goiter is present.

In order to examine the thyroid, a few principles should be kept in mind. Palpation can be done from behind or in front of the patient and the patient’s neck. It is best to examine the thyroid with the neck slightly flexed (relaxing the sternocleidomastoid). When examining the thyroid, note the thyroid’s size, consistency, and texture. Look for tracheal deviation and cervical lymphadenopathy. Ask the patient to swallow and remember that the thyroid is attached to the trachea. During swallowing, both will move upward. This helps to distinguish the thyroid from other neck masses, which will tend NOT to move with swallowing.

Identifying a Goiter
How do you know if a thyroid is ENLARGED? Use the “rule of thumb”: if a lateral lobe is larger than the distal thumb, it is enlarged.

Read more »

Bedside Rounds: How Do You Diagnose and Treat Diabetic Neuropathy

October 3, 2007

Feet 2Commentary by Judith Brenner MD, Associate Program Director, NYU Internal Medicine Residency Program

Diabetic neuropathy is one of the most commonly encountered complications of diabetes mellitus. It is seen in up to 20% of diabetics. Patients typically present with neuropathic pain in a “glove and stocking” distribution with the earliest signs in the feet. Night time complaints of “my feet are on fire” are common. Relying on a patient’s complaint of “pain” or “numbness” is inadequate in the diagnosis of peripheral neuropathy since almost half of patients with ulceration lack these complaints.

To briefly review, physical examination centers on assessment of vibration since it is often one of the earliest neurological signs. A 128-Hz tuning fork is applied to the bony prominence on the dorsal aspect of the first toe just proximal to the nail bed, over the interphalangeal joint. The patient is asked what he or she feels. If impairment exists at this distal bony prominence, the examiner can proceed to test more proximal prominences (such as the malleollus, patella, etc) to assess the extent. Testing the Achilles reflex is also useful. In Sapira’s text, “Art and Science of Bedside Diagnosis”, 15% of patients lack an Achilles reflex. Of those 15%, almost 80% have diabetes. The best technique for eliciting the reflex requires that the patient sit comfortably with legs dangling over the table. The examiner dorsiflexes the ankle and then taps the tendon. Dorsiflexion adds tone to the system, thus accentuating the reflex.

Today, the monofilament test is used most often because of its operating characteristics. It has been found that the inability to sense the monofilament is predictive of ulceration (LR of 2.9). Conversely, the presence of sensation argues against subsequent amputation.

As for treatment, it was first shown in DCCT in 1993(1) that intensive insulin therapy in type I DM reduced the risk of developing clinical neuropathy by 61%. The treatment goal is to slow the progression of neuropathy rather than to overcome pain. Beyond glucose control, treatment has traditionally included antidepressants and anticonvulsants. In a recent systematic review, Wong et al (2) investigated the analgesic efficacy of several different classes of analgesics in managing neuropathy. Studies included in this review included adult patients over 18 years of age who were treated with oral or topical analgesics for short periods of time (2 weeks- 6 months). The drugs investigated included antidepressants (TCAs and SSRIs), opioids, tramadol, capsaicin, anticonvulsants, and NMDA antagonists. In each study, the investigated drug was compared to a placebo. The primary outcome that the investigators were interested in was a 50% reduction of pain, which translated into a “moderate” improvement in global pain. The secondary outcome was the number of patients who withdrew secondary to adverse events.

Read more »