What it means to be a clinician?

What it means to be a clinician?


Vijay Gopichandran


As a teacher in a medical college in the city, I am confronted by a dilemma on a daily basis. The question is what does it mean to be a clinician today? I am not actively practicing clinician. I volunteer my time once a week delivering clinical services in a rural clinic. I work in a resource-poor setting, with a serious financial crisis. There are very limited infrastructure and facilities available in the clinic. Therefore, the medicine I practice there largely depends on careful listening to the patients’ stories, a thorough and astute clinical examination and a lot of instinctive processes. This is in stark contrast to the health system in which I function during the weekdays. The doctors practicing here and the students learning in the hospital have access to cutting edge technology and a good amount of resources at their disposal to help the patients. This leaves me with the question – “Today, what does it really mean to be a clinician?” Is a clinician one who has sharp clinical skills of diagnosis not dependent on technology? Or, is a clinician one who can effectively use technology to help her patients? There is probably no reason to think that a clinician is either this or that. A clinician is probably a combination of both.


Last week we saw a woman in the rural clinic with fever, cough and breathing difficulty. This is a common occurrence during the rainy seasons of Oct-Nov in the Chennai-Kancheepuram area. People who have asthma, develop breathing difficulty and wheezing during this period and require a lot of support in the form of nebulizations, injectable medications, and tablets. On examination, this lady had a fever and a combination of crepitations and wheeze. Crepitations are a type of sound that one hears in the chest when placing the stethoscope on the chest which sounds like the crackling of dry leaves when walking on them, or like blowing air bubbles using a straw in a glass of water. It indicates that there is some fluid collection in the lungs. Sometimes, it is because of heart failure and waterlogging in the lungs and at other times it is because of the collection of fluids in the lungs due to infections such as pneumonia. Wheeze is another distinct sound which one hears when the medium-sized air tubes are constricted or blocked because of asthma. We were expecting this lady to have predominantly wheezing, but she also had these crepitations. So, we were confronted with a clinical picture of a combination of exacerbation of asthma with some form of fluid collection in the lungs. The presence of fever was indicating that there could be some infection in the lungs also.


This is a typical example of the dilemma that I introduced in the beginning of this blog. In the high resource setting in my weekday workplace, I would probably order an X-ray of the chest to see if there is an infection in the lungs. I would also probably get a sputum examination to look for infections and maybe even spirometry (a kind of lung function test) to assess the severity of asthma. However, in the resource-poor rural clinic, the nearest X-ray was 10 km away and is expensive. The nearest government hospital which could offer a free X-ray of the chest was 30 km away. So, we could not do any of these investigations to support our diagnosis.


We did a systematic clinical reasoning process. A young woman, who is a known seasonal wheezer, has presented with fever, cough and breathing difficulty. She clearly says that the seasonal wheezing that she has is different and this time she is feeling worse. She is clinically not stable. She is running a high fever, her heart rate and breathing rate are high. Her chest has both wheezing and crepitations. Given the background history of wheezing, this is probably an exacerbation of asthma. But given that this time the wheezing is not like the routine seasonal one and also given that she has all features suggestive of an infection, it is probably an exacerbation of asthma precipitated by a lung infection. Based on this clinical assessment we started her on nebulization for relieving her asthma exacerbation, gave her injections for the same and started her on oral antibiotics to treat for the infection.


But I was not very comfortable with this decision. Have I initiated an unnecessary antibiotic? In today’s era of emerging anti-microbial resistance (a condition where bacteria develop resistance to antibiotics because of irrational and indiscriminate use), have I contributed to irrational antibiotic use? Am I justified in concluding clinically that this lady had an asthma exacerbation along with a lung infection? These questions were bothering me a lot. All week, I have been reading on the accuracy of clinical examination and clinical signs and symptoms in establishing the diagnosis of conditions. What I found is interesting and I am sharing it in this blog.


In the clinical postings during our MBBS training, we are all taught the important clinical symptoms of tuberculosis. Every medical graduate from India knows these clinical symptoms by heart! These are the combination of prolonged cough, weight loss, fever and night sweats. I came across a study that looked at the accuracy of these clinical symptoms in diagnosing tuberculosis among patients living with HIV/AIDS. We know that tuberculosis is common among people living with HIV/AIDS. The study reported that these clinical symptoms have a sensitivity of 51% in diagnosing tuberculosis among patients who are taking treatment for HIV. What does this mean? This means that by just asking these 4 questions – do you have a cough? Do you have weight loss? Do you have a fever? Do you have night sweats?  - one could pick up 51% of patients who have tuberculosis. The remaining 49% of patients would be missed – also known as false negatives. When the sensitivity is so low (anything less than 80% is considered as low), it means the diagnosis method is missing a lot of patients. The study further showed that when a simple chest x-ray was added to these 4 questions and found to be abnormal, the sensitivity increased to 84%. This means that asking these four questions, doing a chest x-ray and finding all of them abnormal will correctly diagnose 84% of all patients who have tuberculosis. It would miss the remaining 16%. This paper can be accessed here. I know that relying purely on clinical signs would not give the most accurate diagnosis for the patients. But this information that I would miss as high as 50% of the patients if I were to purely rely on the clinical process was disappointing to me. I explored further.


There was a study on the accuracy of respiratory symptoms and signs in the diagnosis of asthma. The study reported that if a patient had the clinical features of wheezing with breathlessness, just this clinical history alone would accurately diagnose asthma with a sensitivity of 82%. To reiterate what this means, among all patients with asthma, if I were to use only the history of breathlessness and the clinical feature of wheezing, I would accurately diagnose 82% of them, but I would miss 18%. This 18 % can only be picked up by lung function tests using spirometry. This is not as dismal a picture as the previous one with tuberculosis. This study can be accessed here.


The other study which I read painted a rather bleak picture of the accuracy of clinical diagnosis. It was a study on the use of clinical features for the diagnosis of pneumonia or lung infection among adults in a primary care setting. This was the study that I was looking for. I thought this study will point towards some answers for my dilemma with the lady whom we had seen in the clinic last week. This study revealed that the positive likelihood ratios of respiratory rate, pulse rate, temperature, crepitations (abnormal crackling chest sounds) were all in the order of 2-3. If a clinical sign is positive, what is the chance that the patient has the disease compared to the patient who doesn’t have the disease? This is the question answered by the likelihood ratio of a positive test. If this likelihood ratio is the order of 10-100, then the test is said to be strongly predictive. However, a likelihood ratio of 2 or 3 is not strong enough to depend on the clinical sign to diagnose the condition. In other words, if before doing this clinical examination the odds of having pneumonia just by history were x, after eliciting a positive sign, the odds increase to 3x. This is not clinically a major leap. This study can be found here.


After reading and reflecting on these studies and the value of the clinical process in diagnosis and delivery of care for patients, I think I have some key conclusions for myself.

1.      Medicine is not a set of algorithms and cook-book recipes. There is a role for history taking, physical examination, clinical reasoning and technological investigations to explore the diagnoses in patients. Even doing all this may not give the accurate full picture. The highest sensitivity and specificity obtained by any diagnostic process is only 99.99999….. and not 100%. 

2.      Relying only on the clinical history and examination to guide the treatment may not be the wisest thing to do. As seen in the paragraphs before, the best accuracy that the clinical symptoms and signs can achieve is 85%. This uncertainty must be borne in mind while making clinical decisions. In other words, all that wheezes is not asthma, and all chest pains are not heart attacks.

3.      The decision on when to rely on clinical process and when to ask for investigations depends largely on the illness, the patient and the context. This why simple decision algorithms cannot function independently of the clinical reasoning process. It is because of this complexity that there is still the role of a medical education which emphasizes the importance of clinical medicine. Because over-reliance on technology without attention to the clinical picture can have its own disadvantages.

I have an answer to my dilemma now. I will continue to teach clinical medicine to the students. The clinical history and examination will help them build the full picture. I will interpret the discouraging statistics on the accuracy of clinical findings differently. These data on poor accuracy does not mean they are not useful. It only means they are not accurate and so cannot be used as sole modalities of arriving at a diagnosis. But the clinical process helps arrive at a complete picture of the patient. It has other very important roles to play in the doctor-patient clinical encounter. It elicits the expectations of the patients from the treatment. It helps the doctor understand what bothers the patient most. In many medical conditions even before arriving at a diagnosis, a thorough understanding of the expectations and concerns of the patients can help the doctor address and relieve these issues. It is for these reasons and for building a background for the diagnosis that clinical skills are still important.