Parsimonious. My teachers taught me this principle at the bedside during my residency. A single diagnosis should explain all symptoms and signs whenever possible.
That was in 1980. Physicians relied on their brains and stethoscopes. They took detailed histories, examined patients thoroughly, ordered few tests, discussed differentials, and took pride in getting the diagnosis right.
I didn’t know then that my teachers were referring to Occam’s razor—a logical principle often applied in medicine. It states that the simplest, single explanation is usually correct. When patients presented with fever, a heart murmur, and an enlarged spleen, we diagnosed infective endocarditis. When alcohol-dependent patients arrived with swollen legs, yellow eyes, and a distended abdomen, we attributed everything to liver cirrhosis.
We rarely considered that a febrile patient with a heart murmur could also have typhoid. Or that a long-term smoker might die from a pulmonary embolism rather than an exacerbation of COPD. Or that an anxious girl gasping in the ER might have diabetic ketoacidosis rather than a psychosomatic disorder.
Yesterday, a medical student reminded me of an important lesson: Never close a differential diagnosis too soon.
For three weeks, she had a blocked, runny nose. ENT specialists diagnosed acute bacterial sinusitis. Low-grade fever, headaches, and yellow nasal discharge fit the picture. She took antibiotics, antihistamines, and decongestants. Her fever subsided—then returned. She felt drained, self-diagnosed allergies, and prepared for her pre-final exams, exhausted and anxious.
Her father, a physician, travelled 20 hours by train to see her. He asked if I could visit her in the guest house where she had temporarily moved from her hostel. I agreed.
She looked weak. I asked pointed questions, traced the illness’s timeline, and noted her fever patterns. I checked her vitals, sinuses, throat, and chest. Her nasal lining was red, her throat congested. “My nose is still blocked,” she said, “and my headaches are terrible.” She blamed allergies. I reassured her. “The infection may be resistant to the antibiotic,” I told her father. I prescribed a new one, hoping it would work.
Her father remained silent. I sensed his unease. He wanted a full workup. I asked her to visit me in the OPD.
The next morning, she came in. I ordered blood tests, malaria screening, a Widal test, and X-rays of her chest and sinuses. I also sent her for another ENT consultation.
Three hours later, her father called. Plasmodium falciparum. Life-threatening malaria.
Relief. Shame. Relief that the lab provided an answer. Shame that it hadn’t come from my own reasoning. I had fixated on sinusitis, distracted by her talk of allergies. I hadn’t considered malaria—despite practicing in Sevagram, where we see a hundred severe malaria cases each season.
Why had I missed it? I had closed my differential diagnosis too early. Occam’s razor had misled me. I assumed one cause explained everything. I hadn’t asked, “What else could this be?”
In How Doctors Think, Jerome Groopman describes why physicians miss diagnoses. Not from ignorance, but from cognitive traps. He offers a simple three-question checklist:
- What else could this be?
- Could two conditions be present at once?
- Is anything inconsistent with my working diagnosis?
Had I asked those questions, malaria would have surfaced sooner.
How do we avoid these errors? Approach each case with fresh eyes. Keep an open mind. Don’t rush to closure. And without ordering unnecessary tests, remember that a simple checklist can prevent a serious mistake.
Your post gives an all time advice to all medical practitioners.- Meera Kulkarni
A valuable advice to all. An open mind to all non responding symptoms definitely help us reach to the root cause.
SP,
In the Introduction sessions, I used to convey to our interns & PG students that they should not over investigate and add unnecessary work load to the lab. I used to quote the following very old BMJ reference ….further convincing them that this statements was still valid.
"The case for logical requesting of an investigation:
Why do I request this test?
What will I look for in the result?
If I find what I am looking for, will it affect my diagnosis?
How will this investigation affect my management of the patient?
Will this investigation ultimately benefit the patient?
Ref: Asher,R (1954) Straight and crooked thinking in medicine, BMJ, 2,460-462".
Now after going through your blog, I am convinced…this concept does not deserve to be emphasized.
– Reddy
Wonderful post Sir, and a very pertinent question asked. This tunnel vision is a global phenomenon. We have actually stopped thinking and forget that in todays era, multiple / complicated/superadded infections are increasing rapidly. We routinely see patients with co-infections of dengue, malaria, leptospira all in one at the same time. The message is clear… keep the mind always open for more…
Dear Kalantri Sir
Sure every one of us has come across such a situation once in a while.
I spent three years in residency trying to solve the puzzle one size to fit all, because frankly that is what is always emphasized. A little deviation for this dictum and comments like, “Do you think this patient is a museum of diseases?" are shot at us. I agree with Reddy Sir, all we have taught the UGs and PGs is asking ourselves before u order…
Reddy sir says that this concept does not deserve to be emphasized anymore. But I disagree on that.
I still remember a patient on Bed 25 when I was a registrar in Unit 1.This patient had multiple problems and we did, as is customary in Unit 1, a battery of rare tests to get to HIFI sounding diagnosis. That patient simply thanked Gupta sir in front of all PG , nurses and others in the ward for letting him know what disease he is gonna die of as he had exhausted all his resources with nothing left for the treatment.
On the other hand, here at CMC, the patients i face don’t let me start chemo unless they are through CTs and PETs and FNAC a numerous times so that every family member is finally convinced that indeed it is Cancer.
Here, I just cannot afford to under investigate.
But that he fun of it all, I feel both ends should be presented to those, we teach and let them choose what they feel is right. It should come as a personal decision to choose the line they would walk.
As regards your case, I always wonder why such adventures always occur in VIPs, Professor, Staff and medical students and never in the thousands of other patients we cater to in the OPD. That's the question to answer…..
Or as you always say, probably it’s the best and worst case bias that we always remember….
Amit Bhatt
Dear Sir
Saprem namaskar
Thank you for your beautiful blog.
Yes, we all should have High Index of Suspicion while dealing with the common problems. Again I would like to say that the uncommon manifestation of a common disease is more common than the common manifestation of an uncommon disease.
Putting a full stop by just saying that it’s viral is not at all beneficial to the patient. When we all know the streptococcal sore throat can very well cause many hearts to fibrillate, why shouldn’t we help the beating hearts to dance in sinus rhythm. We should always provide the maximum benefits to the ailing patients. Let us not restrict ourselves to penidure , Lasix & digoxin ; this is the time to go for BMV at least.
Although we follow the clinical governance and the EBM in true sense, let us have due consideration for the common things. Evaluate the case with open mind with most specific investigations and arrive at a specific diagnosis.
Apart from the presenting problems, the other co morbidities must be given due consideration and should be investigated. When we are dealing with fever, atleast Malaria,Meningitis,Typhoid,Tuberculosis,Peumonias,Hepatitis & UTI should be thought for and also keep an eye over other inflammatory and malignant possibilities. But generally we don’t think what we don’t come across. I still remember one more case of ’95 batch student who had chest discomfort and finally found to have DVT with pulmonary infarct. I personally suffered very badly during MBBS days of malaria simply because PS was negative and even the C/Q was not considered. So try to realise the pain & suffering by the patients and have due considerations for them.
But all this is possible when there is a expandable space inside ( Between two ear pieces of a stethoscope) !!!
The intelligent horses can lag behind the experienced donkeys; yes the past experience taught us a lot and creates alerts for the future. So keep eyes wide open before putting a full stop. Intelligence and experience both hand in hand can show us a bright way ahead.
Thank you & best regards
Very good post Sir,
We commenly are surprised with the investigation reports which do not go hand in hand with our clinical diagnosis.
I strongly beleive that any fever of more than 5 days duration needs to be throughly evaluated,whatever may be the clinical judgement .
I agree to Dr Dhananjay that putting a fullstop by saying ITS VIRAL is not beneficial for the patients
Thanks for sharing this.
Regards
Mugdha
Dear Sir,
Excellent post.
You talked about Occam's Razor. Even in scientific principles it is still true. Given the history and presentation, no doubt sinusitis was the prime suspect. However as you mentioned we would see probably 100+ malaria in a single season, probability of Occam’s razor for malaria is equally on. Occam's Razor – All things being equal the simplest explanation tends to be true. In our case with endemicity of Malaria for example, all things were not equal and Occam's Razor did not apply.
The case here is our fixation for what seem to be obvious. I suppose it is process of mind to prioritize the issue. Here few years back there was case where Anesthetist could not intubate patient after induction of Anaesthesia. He and his colleague so fixated on getting airway that ended up ignoring the ventilation part of it. The sequence of events then leads to neurological insult. The case went on as classical case for what is dubbed "Human Factors". In our case as well it is this part of Human Factor that seems to be at work. However given the history, when we saw our prime suspect of Acute sinusitis did not show response to the usual treatment, we fixed ourselves to why the treatment was not effective and attempted to change the treatment. Our focus changed from the diagnosis to treatment response. This was the time when a red flag should come up – usual diagnosis, usual treatment, no response, why not? Is our frame of reference wrong i.e. diagnosis? What logic demands at this time is next dimension of thinking. Multilevel thinking simultaneously? Given the complexity of today’s life style, co-existing ailments are perhaps more than reality. e.g. The list of differential diagnosis of Acute Appendicitis runs in 20s. And it would be possible to some of them co-exist.
In cardiovascular cases for example history and physical examination would point to the diagnosis with few differentials even auscultation would be used only to confirm or refute the differential. Investigations come later. However that is not the scene at present at least here. A bias of referral letter from GP, Pediatrician to start with. Time in clinic and first thing many cardiologists do is to put echo probe on chest. I suppose this is other extreme.
However, when you look at resources, limiting investigations makes sense. Something called routine has to be evaluated in the context of will we be wiser when the result come up. If answer is No, then the investigation is not required. One of the surgeons I worked in Mumbai would order pre op or post op investigations quite infrequently and would write urgent even more rarely. All the labs and technicians knew that. Once he asked me to arrange for some radiological investigation. Being new house office, I was preparing myself for battle with radiologist to get it accepted. To my surprise radiologist did not even ask me why and what. His remarks were if this surgeon has asked something it must be necessary.
But this is what is the difference between a machine and human. An exercise in multilevel and lateral thinking which I suppose is impossible for a machine to compute. e.g. Nowadays ECG machine gives you the possible diagnosis with the ECG strip. But there is always an asterix * marked with that – * Unconfirmed. Here comes the human multilevel thinking and as part and parcel the pit fall of Human Factors.
Deepesh Trivedi
Dear SP,
Congratulations on writing this nice article which should encourage physicians to be more alert, methodical, patient and also approach the every case with flexible, broad & rational mind. Pattern of diseases, virulence of germs & clinical presentation also tends to change with time. Every case is teacher in itself. Hence we continue to be students life long.You have done a good Job indeed!
-Vivek Poflee
As usual you have explained and taught again! Do keep sharing ur thoughts which are pearls for me/us and helps a lot in practice and life…
Mohd Abdul Sami
Dear Kalantri sir,
The post was superb. It questions the common belief that experience raises our clinical skills and after practicing for so many years diagnosis of commoner ailments becomes easy for us. I think failure sometimes is because of our experience only. Actually every patient is unique in itself and so commonsense developed by our experience doesn't work always.
I am reading "The Checklist Manifesto" by Dr Atul Gawande these days which deals with the same question posed by you. If you haven't gone through it, I heavily recommend it for every seasoned clinician.
Anupama Gupta
We all learn from our mistakes. Jerome Groopman's three questions seem to be a nice guide to apply in problem cases which we all should remember.
Very well written indeed.
Pushpa Chaturvedi
The more time I spend outside medicine peering in, the more I marvel at the ability of diagnoses being made on pattern recognition and being right most of the time.
There used to be a time when medical information, investigations was available to only those who went through years of passing through the rigors of a medical education. Nowadays, I meet scientists, teachers who probably know more than a 3rd year resident because they have just looked it up on uptodate or Merck manual online.
So, is the value of a trained physician simply to know when to order what test and then to be a trained interpreter of results and of course, an excellent communicator of these issues?
Anyway, fever at MGIMS, hmmm where have I heard that before…seems like a bit of a jinx!!
Dear Sir
I do agree with what one of the persons above said "this happens only in VIPs" as we are overcatious and tend to over diagnose. As surgeons when we operate a staff or Doctors or VIPs somer times even after utmost care we tend to get smiling wounds on stich removal and this is what is VIP Syndrome. and to say Doctors are worst patients. In most incidences we tend to forget commonesst things very commonly when it comes to VIPs
Dilip Gupta
Respected Sir,
Thank you for the valuable post.
It was really an enlightening one. Ocam's Razor is the the one which we have always been following but its very true…. that it cant always be true.
After making the differential diagnosis we just keep on ruling out the most common ones and just tend to forget or underestimate the rarer or the less common ones. This incidence makes us realise that in this Dynamic world with ever changing environment , climates , bugs , the presentations of the diseases , the symptomatology is definitely getting affected as well. This experience really makes us understand the need to make a wide clinical diagnosis and investigate throughly till we get the answers…..
Regards
Swati
Dear Sir,
Excellent post.
You talked about Occam's Razor. Even in scientific principles it is still true. Given the history and presentation, no doubt sinusitis was the prime suspect. However as you mentioned we would see probably 100+ malaria in a single season, probability of occam's razor for malaria is equally on. Occam's Razor – All things being equal the simplest explaination tends to be true. In our case with endemicity of Malaria for example, all things were not equal and Occam's Razor did not apply.
The case here is our fixation for what seem to be obvious. I suppose it is process of mind to priorarize the issue. Here few years back there was case where Anaesthetist could not intubate patient after induction of Anaesthesia. He and his colleague so fixated on getting airway that ended up ignoring the ventilation part of it. The sequence of events then lead to neurological insult. The case went on as classical case for what is dubbed "Human Factors". In our case as well it is this part of Human Factor that seems to be at work. However given the history, when we saw our prime suspect of Acute sinusitis did not show response to the usual treatment, we fixed ourselves to why the treatment was not effective and attempted to change the treatment. Our focus changed from the diagnosis to treatment response. This was the time when a red flag should come up – usual diagnosis, usual treatment, no response, why not ? Is our frame of reference wrong i.e. diagnosis ? What logic demands at this time is next dimension of thinking. Multilevel thinking simultaneously ? Given the complexity of todays life style, co-existing ailments are perhaps more than reality. e.g The list of differential diagnosis of Acute Appendicitis run in 20s. And it would be possible to some of them co-exist.
In cardiovascular cases for example history and physical examination would point to the diagnosis with few differential even auscultation would be used only to confirm or refutr the differential. Investigations come later. However that is not the scene at present at least here. A bias of referral letter from GP, Paediatrician to start with. Time in clinic and first thing many cardiologist do is to put echo probe on chest. I suppose this is other extreme.
However when you look at resources, limiting investigations makes sense. Something called routine has to be evaluated in the context of will we be wiser when the result come up. If answer is No, then the investigation is not required. One of the surgeon I worked in Mumbai would order pre op or post op investigations quite infrequently and would write urgent even more rarely. All the labs and tecnicians knew that. Once he asked me to arrange for some radiological investigation. Being new house office, I was preparing myself for battle with radiologist to get it accepted. To my surprise radiologist did not even ask me why and what. His remarks were if this surgeon has asked something it must be neccessary.
But this is waht is the difference between a machine and human. An exercise in multilevel and lateral thinking which I suppose is impossible for a machine to compute. e.g. Nowadays ECG machine gives you the possible diagnosis with the ECG strip. But there is always an asterix * marked with that – * Unconfirmed. Here comes the human multilevel thinking and as part and parcel the pit fall of Human Factors.
I tried to post it on the blog forum earlier, but some how it would not, human factor (error at least in my case). but I seem to have overcome that today.
With warm regards.
Dipesh
Dear SP,
Everything changes with time,including perception.The availability of such large number of investigations have added to the woes of physicians.But all said and done,it has its uses.One of my friends had PUO .Urine exmn pointed to UTI ,which was treated.A positive widal resulted in plethora of antibiotics.USG was also WNL.He consulted a Urologist and a nephrologist as well.A sudden fall in Hb was like an alarmbell.Pt was shifted to Medanta Medicity where a whole body scan lead to a diagnosis of lymphoma,with large lymph nodes pressing kidney.Pt has recd full course of chemo and is recovering.
There ends the story of ideal tests and cost saving!
VK
Dear Sir,
An excellent post. It makes me continue to appreciate the breadth of medicine, the perspective of which it is not very difficult to lose, once one goes into a specialty to know more and more about less and less!
One of the reasons I continue to still do- and enjoy- general medicine on-calls is that patients come with symptoms and not necessarily a diagnosis.
Another thought which I have to keep reminding myself- and was brought home again reading your post- is that specialty induced blinkering of vision, while in some case necessary, is not necessarily always good. When I read about sinusitis, one of the differentials that came to my mind (relatively rare though it is, but not uncommonly seen in Rheumatology) is Wegener's!
Many thanks again for sending me the link to the post, and I will continue to look forward to receiving more.
Regards,
Mihir.
Dear Sir,
It was a pleasure reading your blog. I totally agree with the statement that we should not close the differential prematurely.
But I do not want this to be misinterpreted as a licence to order all the investigations available, a system that is plaguing the US healthcare system.
These missed diagnoses are an exception to the rule. Taking your case as an example, I would think of sinusitis as my diagnosis. But only after the patient has not responded to standard sinusitis treatment would I rethink my differential and go over the entire checklist and order specific investigations to narrow my new differential.
On an average 20% of reported lab values are erroneous. So a comprehensive history and thinking of first things first would be the way to go.
I still believe that the Occum's razor be taught to the medical students. Or else there is no limit to what tests can be ordered. There is no substitute to history taking and a careful examination followed by directed lab investigations. These are the teachings of Sewagram that hold true to date. They have helped me wherever I have been.