Uncategorized

How the Rest of the World Does It: New Delhi’s Safdarjung Hospital

I visited Safdarjung Hospital in New Delhi today – an institution with  1,531 beds and 145% occupancy rate.  Yes, 145%.  You do the math.  A lot of bed sharing and asking families to bring in cots.  It’s right across the street from the All India Institute of Medical Sciences (AIIMS), the premier public healthcare institution in India.  While both AIIMS and Safdarjung are run by the federal government, only AIIMS is renowned for famous specialists, world class facilities, and an international reputation to boot.  Safdarjung doesn’t suffer such burdens – its specialists are not well known, facilities are dilapidated, and you probably have never heard of it.

I spent several hours walking around, talking to lots of physicians, visiting ICUs and cath labs.  I visited the outpatient department where 7,000 people show up every day, many lining up the night before, to get a ticket by 11 a.m., when registration closes and those who haven’t gotten a ticket are out of luck.  In the ER, there was a line of between 50 and 100 people waiting to get rabies shots.  This is the hospital where every poor person in Delhi unfortunate enough to get a dog bite is sent.  They have the rabies serum.  Most other public hospitals do not.

Safdarjung has “efficiency” baked in.  In a typical year, they do 800 cardiac surgeries, 2,000 angioplasties, 3,000 echocardiograms, and 100,000 EKGs.  They see tens of thousands of patients in the cardiology clinic.  They have 4 (yes, four) full-time cardiologists on staff.  The rest of the work is done by medical residents, who call when they get into trouble.  Brigham and Women’s Hospital, which probably doesn’t have one quarter the volume of this place, has 140 cardiologists.  The patients at Safdarjung pay essentially nothing.  Even their medications are free.  For those who are not extremely poor (and I doubt there are many non-poor patients who go to Safdarjung), you do have to pay for your own devices.  Need a stent?  Bare metal ones cost $200 to $1000.  Drug eluting stents are $1500 to $2500.  You get to decide which one you want.  They have a chart with pictures and prices that looks a lot like a dinner menu. 

What is remarkable about Safdarjung, though, is not its bustling hallways and jam-packed ER.  It’s how well it seems to work.  I visited a large ICU with lots of patients on ventilators, and a single medical resident running the place.  During the time we talked, he scanned the room and gave out orders. Everything seemed under control.  If you believe in the data on the volume-outcome relationship (and you should), it’s clear why this place claims to have terrific outcomes. They very well might.  Yet, as I walked around with the chief of cardiology, I asked him about their cardiac surgery mortality rates.  He assured me mortality was low, “comparable to international standards” (whatever that is).  I pushed him – he said very few patients died after procedures.  When I pushed a little more, he got annoyed, wondering if I was accusing him of running a poor quality hospital.  I backed off.

This is a place that seems to have no time for data.  At each step, I asked if they tracked outcomes.  They didn’t.  They know the latest evidence.  They could easily tell you all the studies that underpin the Hospital Compare quality measures and assured me they did all of those things “always”.  Patients always got antibiotics quickly.  Thrombolysis or primary PCI was never delayed.  No one went home without a beta-blocker.  Yet, several clinicians seemed to grow tired when I asked gently if they tracked their data.  They didn’t.

Safdarjung hospital is a marvel.  It has huge volumes and clinicians who are clearly both incredibly talented and dedicated.  If I were a guessing person, I’d say it probably achieves 80% of the quality of U.S. hospitals at 10% of the cost.  However, while I’m confident on the cost, I’m guessing on the quality.  For some procedures, they probably do better than the average U.S. hospital.  The upside of working in a country where high volume is easy to achieve. Of course, Safdarjung does it with none of the creature comforts we’d want in a hospital (think 145% occupancy rate and patients having to double up).  In the cardiac unit, there’s one monitor for every two patients, and they switch off depending on who is sicker at any given moment.  In the U.S., we obsess if it’s OK for a patient to take off their cardiac monitor for 10 minutes to take a shower.  This is how Safdarjung does so much with so little.  Their clinicians “cut corners” we are not willing to cut.  Its not clear to me that they are practicing worse medicine than I am.  The corners they cut are often of little or no clinical consequence.

Safdarjung is not a place that would score very highly on “quality culture” surveys.  It’s focused on efficiency in a way that few places are, and it probably has no choice but to prioritize this (think 7000 outpatient visits and 400 inpatient hospitalizations every day).  It’s overwhelming.  But, if Safdarjung could put in real metrics for clinical quality, pick some low hanging fruit (a simple EHR would be really helpful) and stop assuming good patient outcomes, it could change the culture of clinical care in India and probably surpass a majority of U.S. hospitals on safety and effectiveness.  Adding private rooms and some decent food would still leave it 85% cheaper than the average U.S. hospital. At what price point would Americans be willing to travel to India for their hip surgery?  I suspect Safdarjung wouldn’t be their destination (it’s too busy caring for the poor to invest in medical tourism).  But, there are lots of hospitals like it in Delhi – high volume centers with an appetite for new technologies and the creature comforts we would want.  Their problem is that they neither reliably measure quality nor make its improvement a priority.  Therefore, we’re left assuming “international standards”.  If there is anything we’ve learned from looking at quality in the U.S., it’s that hospitals across the street from each other can have profoundly different outcomes.  There is, unfortunately, still no international standard.

Safdarjung is an amazing place:  a high volume, efficiency-driven institution that seems to deliver pretty good care.  If it could just sprinkle in some quality measurement and make quality improvement a routine part of how it delivers care, it would likely have a profound effect on how hospital care is delivered across India and beyond.  Who knows, it might even have something to teach U.S. hospitals.

Ashish Jha, MD, MPH is the C. Boyden Gray Associate Professor of Health Policy and Management at the Harvard School of Public Health. He blogs at An Ounce of Evidence.

17 replies »

  1. This was an amazing read. I was told by a colleague about the way Indian hospitals are run and how quickly people are able to be treated by quality medical staff – in comparison to Australia. In fact, I have a friend that travels to India for medical treatment (he is an Indian), especially dentistry, and is always thrilled with the treatment, price and efficiency. Thanks for sharing.

  2. Nice article Ashish. I worked as a medicine resident in Safdarjung for 3 years and then did residency is US as well. I agree that they are well organized and workload is high to the extent that I have alone done 15 admissions and seen at least 50 patients in the ER on the same day. But I feel there were some people who could be saved died, so critical care part is very deficient. ICU has just 10 beds-is that enough for the number of people who walk in-obviously not. I somehow felt that at least 80% of people who survived the hospitalization were not very sick to begin with. You will never know that the stenting the cardiologists are doing is appropriate or not. Again, as you appropriately pointed, there is no data on as to how many people die after their cardiovascular (especially valvular surgeries). I will take my family there for any vaccinations and antibiotics but that will be it. I think that there obstetrics department is fine as well.

  3. Dear Sir,

    Why the community wants a social service from Doctors and in return Doctors are deprived of even basic security. Why no one talks about engineers or IT people. Why it is a crime for a doctor to think about himself.

    If the working conditions will improve and the Doctors will get even half of what they deserve , I am sure you will see a drastic reduction in the brain drain and improvement in the Indian Health system as well.

  4. Dr Ashish has given a very fair and unbiased description of Safdarjung Hospital. I agree with him , that with small care and arrangement , we should collect data for evaluations and other important purpose . In India 90% Govt. Hospital are giving services , more than their capacity , although lack of funds,space etc is leading to many problems and so called 1/2 star rating . He must have noticed , poor ratio of Patient and paramedical staff also.
    This can further improve with more funds & more Doctors & paramedicals.
    Unfortunately , so many doctor ,who born and studied here , go abroad for better prospects , reducing No of professionals , our country is producing . After all person advantages come first..

  5. Not many people realise the quantity of health care delivery being delivered by such hospitals .

  6. all i cn say…. a nice scanning of our safdarjung hsptl by u……….bt personally i feel , li’l bit more observation required , of different deptt to get the whole idea bout this hsptl……its really really efficient within its limitations of resources,,,,,,,,,,,,,,m enthusiasitic being a part of this hsptl…….

  7. This is such an excellent, matter of fact and beautiful observation by you, Dr Jha, of Safdarjung hospital……. I work here as a faculty in microbiology….it has fulfilled a large part of my dreams as a medical microbiologist and a teacher through opportunity to interact with excellent students and as a clinician I couldn’t have asked for a better situation to understand infectious diseases and support the free healthcare system of our country …….am very proud of all what you say……thanks……hope this is read widely

  8. Wel said sir,safdarjung hospital in spite of all d odds,caters to more than 5000 patients which is not an easy thing.thank u for highliting this and hope our government wil make efforts to further improve its quality and standards.thanks once again

  9. First of all, I appreciate your effort in writing such a nice blog. I have been working and residing at Safdarjang Hospital for last 13 years now and I must say it has become more than a home. I had an interaction with you at the canteen and good to see that you came out with such fine details.

    Safdarjang was started as an American Army Base Hospital during World War ll and you can still see the old barracks across the ring road. It is still functioning smoothly and growing steadily for last 70 years.
    Its a tertiary care hospital and receive referred patients from different hospitals of whole Northern India. It is very difficult to maintain top quality with such a high quantity. I think, to compare the two hospital systems (India & US) would not be fair.

    The patients we receive here have already taken treatment in multiple hospitals before reaching here and usually are in the advanced stages or terminally ill. With the improvement in quality over the last decade has now diverted a lot of upper class patients as well to Safdarjang Hospital and with the new Multi Superspecilty block with 500 beds coming in next 3 years will add further more.

    I disscussed this quality issue a lot of times with my friends but as far as Safdarjang is providing free treatment to every individual irrespective of class, religion, locality or nationality (which is the best part about Safdarjang) it will never reach the quality standards of Corporate or US hospitals that have much less patient load.

  10. I was recently a patient at Samsung medical center in Seoul – I was highly impressed by the quality of care, efficiency of care. Despite the size of this hospital, the patient does not feel lost.

    For those comparing hospital and medical systems, this is worth a look.

  11. Good points, rbaez.

    One tough issue for America is that if we ever want cheaper hospitals, I think there will have to be a lot of layoffs.

    When you look at how hospital employment has grown in the US during the last 15 years of reduced utilization, it is just astounding. Hospitals have been a New Deal all to themselves in terms of a unacknowledged national jobs program.

  12. Margalit,
    “I wonder how the Observer Effect of measuring every little thing, documenting hundreds of data points that have nothing to do with patient care, and trying to manipulate the “data” and the workflows so it all looks good on “paper”, would affect the performance and efficiency of this hospital.”
    I think that’s an excellent point. A middle of the road example between India and the US would be German hospitals, that deliver reasonable care but cut more corners than US hospitals (the German ones are usually well maintained, often shiny facilities with much leaner staffing and on average lower acuity and longer stay). One example: in Germany, most common medications were stored (at least a good decade ago, I don’t think that has changed) on the ward, and could simply be given by the RN on written or verbal order. In the US hospital, you have the benefit of a central pharmacy looking at the order (i.e. checking for allergies, interactions) and sending it up, but this causes a lot of additional work/cost and produces delay (which at times may be relevant in terms of patient satisfaction and medical quality). Another relevant example would be that residents are on average less supervised and major workhorses of many institutions down to community hospitals; there is certainly a little bit more room for error there … but is it really relevant in the big picture ? The big question by m is: would you go there yourself, or send a family member? I have a family member who just underwent a coronary angio and will likely have heart valve surgery … I think – looking at Emails and d/c summaries from her docs – that her care at middle of the road institutions (i.e. outpatient care and a busy community hospital) is – so far – very good.

  13. I wonder how the Observer Effect of measuring every little thing, documenting hundreds of data points that have nothing to do with patient care, and trying to manipulate the “data” and the workflows so it all looks good on “paper”, would affect the performance and efficiency of this hospital.
    It seems to me that the better idea would be to build another one, or expand this one, so the occupancy rate is more manageable, and let those obviously excellent physicians do what they were trained to do and what they are passionate about doing, without having to cut too many corners.

  14. Fantastic post!

    I hope that John Goodman is reading this. He will see the upside and the downsides of medical tourism.

    Our solution? Give American hospitals about 65% of their current budgets in global lump sums, and let them come to their own solutions. The persons who run American hospitals all have very high IQ’s, whether they are MD’s or have a master’s in administration.

    Salaried cardiogists and nurses making $125,000 a year will have a problem in global budgeting. We will survive.

  15. Ashish, I really appreciate the spirit of what you are saying. I am a physician, am Indian and have worked in quality consulting. I think that the poorer, high volume hospitals in India are absolutely amazing in what they are able to accomplish. That being said, the question I would always ask is would I go there or want a family member to go there. And I know the answer would be no.

    The problem with the care at these facilities is, as you said, the corners they are willing to cut. For the most part, maybe people won’t die. But some will when they shouldn’t have. Some will get bad infections when they shouldn’t have. Some will have poor outcomes when they shouldn’t have.

    But in America, we have the opposite problem. The level of expectation of perfect outcomes has become so high that the cost for each patient and for the system overall has become untenable. Somewhere in between the two is probably where our answers lie. I think looking at AIIMS, Wockhardt and similar large volume, high quality hospitals may help us to have some of the solutions you are trying for.