Hospitalist vs Primary Care 30 Day Discharge Cost Utilization Study.

Here's the latest study comparing cost utilization of hospitalists vs primary care in the 30 day post hospital discharge period of time.  It's titled Association of Hospitalist Care With Medical Utilization After Discharge:  Evidence of Cost Shift From a Cohort Study.  I had an opportunity to read the study in full.  I called my hospital library and had the pdf article in my email in less than 24 hours.  Now, that's amazing.  Thanks to the education staff for a job well done.  Presented in the August 2nd, 2011 edition of the journal Annals of Internal Medicine, Volume 155 Number 3 Page 152-159, the study concludes that decreased length of stay and hospital costs associated with hospitalist care are offset by higher medical utilization and costs after discharge.

In summary, hosptitalist patients had an adjusted length of stay 0.64 days shorter and $282 less than patients cared for by primary care physicians, but total 30 day post discharge costs were $332 higher.  These additional charges were defined as 59% from rehospitalization, 19% from skilled-nursing facilities, and 22% from professional and other services.  OK fair enough.  Let's come to that conclusion.  Let's say Medicare pays more in the 30 days after discharge for patients cared for by hospitalists than they do for patients cared for by primary care physicians.  Is it possible that hospitalists shift cost to outside the hospital?  Yes.  It's possible.    Let's just assume that hospitalist patients are equal in all ways to patients cared for by primary care physicians.   It's also possible that primary care physicians shift cost into the hospital where DRG payments don't account for or fund for additional service.  In an environment where hospital Medicare profit margins have been negative for almost a decade, doctors who try and get everything done in the hospital aren't doing their community hospital  any service.  A hospital that goes under will provide no care for anyone.  

The study concludes that Medicare pays an additional 1.1 billion dollars a year for patients cared for by hospitalists compared with primary care physicians in the 30 days after discharge.  It's quite possible that this additional money represents underpayments to hospitals using underfunded, negative profit margin DRG payment rates and should be provided to hospitals as a matter of sound economic principles.  Perhaps hospitalists are the ones who realize that Medicare DRG payment rates do not account for the all you can eat hospital buffet some patients and doctors expect in the hospital.  But I don't believe either to be the case.  The study admits, as an observational study, it is subject to selection bias.  That means our patients may not be an apples-to-apples comparison with primary care physician patients.   Here are some reasons to question the data
  • The data is from 2001-2006.  A lot has happened, especially in IT, in the last 10 years.  The discharge process has improved in many ways, and we know that a healthy discharge process helps prevent readmissions more than any other aspect of care.  It is quite possible that the 2011 discharge process has  eliminated or reduced many readmissions that would have otherwise occurred in 2001, when resume home meds was the most common acceptable discharge process.
  • Some key differences in the patient population studied
    • Patients cared for by hospitalists were defined as low income at a rate of 21% vs 18% for the PCP.  I think, in general, the poorer you are, the sicker you are.  There is also a direct correlation with smoking and being poor.  So, I'd like to know if hospitalist patients had a higher rate of ongoing or previous tobacco abuse.  We now that quitting smoking after a heart attack is equal or better than any pill.  We also know that half the people start smoking within 20 days of discharge and that increases the risk dramatically of another devastating event in the post hospital discharge period.  Unfortunately, the study did not discuss smoking status at all.
    • Weekend admissions accounted for 29.2% of hospitalist cared for patients, but only 24.6% of primary care physicians.  Why is that?  Many PCPs don't work weekends.  Many hospital services are not available or are delayed on weekends and will  be deferred to the outpatient.  
    • For hospitalists, 8.5% of patients were admitted from nursing homes, while only 4.9% of PCP patients came from that setting.  This part is huge.  Why?  Patients are in a nursing home because they need help.  Their bodies are failing.  Their minds are failing.  They present with a protoplasm that places them at a higher risk of 30 day death, disability and bounce back. The nursing home-hospital merry-go-round is a well documented phenomena in hospitalist medicine. They are the TPN (trach-peg-nursing home) patients of the Medicare National Bank. They will be in and out of our hospital until they either die or their family decides enough is enough.  That's where an excellent palliative care team can stop the madness.  Does this characteristic explain the findings of higher post hospital discharge costs?  I don't know because the study did not attempt to define it further.  My guess is yes.  It just takes a few smokers and nursing home patients to really skew the data.  These are the people that bounce back.  
    • The mean number of comorbid conditions for hospitalist cared for patients was 3.3, compared with 3.1 for PCPs, indicating a patient population that was more complex, despite a younger age at 77.3 years old versus 78 years old for the primary care doctors.   Also, for patients in the 75th percentile, hospitalist patients averaged 5 comorbid conditions vs just 4 for the PCP.  I can't even imagine what the 95th percentile is, as my experience tells me that the super complicated patients are all deferred to the expertise of the hospitalist.
    • Hospitalist cared for patients had more outpatient physician visits in the year preceding hospital admission (11.7 vs 11.4) indicating a patient population that required a higher intensity of service (likely with multiple subspecialty services).  With the 25th percentile of PCP patients seeing a physician five times a year and hospitalist patients seeing an outpatient six times in a year, and a median of 9 vs 10 respectively, it seems to me that hospitalist patient need more help all the time.   That means they are going to need more help after discharge and they represent a higher risk of bouncing back to the hospital.  
    • Hospitalist cared for patients came from metropolitan areas greater than one million people at a rate of 54.4% while only 50.4% for PCPs.  There are more services available for outpatient management in a city of one million than there are in a city of 100K.  They are going to cost more after discharge.  And because they need more appointments for their higher number of medical problems, they are more likely to miss necessary care.  Because they are poorer and are more likely to smoke, they may not have money for gas in their car to make their appointment to the cardiologist, nephrologist and pulmonologist.   
    • Patients cared for by PCPs  had more visits to their PCP and fewer visits to other outpatient providers and fewer ER visits as well after hospital discharge.  It's quite possible that PCP selection bias is also at work here.  Daily, our hospitalist group works with primary care doctors who pick and choose who they desire to see in the hospital.  They know their patients well.  They also know which patients they would rather have a hospitalist care for.  In fact, this is openly admitted and often painfully laughed about in direct communication.    Often, there will be two patients in the emergency department, at the same time, with the same primary care doctor.  They will take one and leave the complicated patient with difficult family and social dynamics to the hospitalist.  I can't see that this study has attempted to define the extent of mental illness or personality disorders in the differentiation of hospitalist vs PCP cared for patients.  This is alikely a big yet under appreciated cost center for Medicare.
    • There is no attempt to define a functional score for either group of patients.  How well do they walk?  Are they falling?  Are they experiencing significant issues with dementia?  Functional capacity is a major driver of resource utilization.  It's quite possible that hospitalist patients require greater post hospital discharge care in a skilled-nursing facility because their lack of  functional capacity requires it.  
    • The path of least resistance from a nursing facility to health care is the ER, not the primary care doctor's office.  You discharge more patients to a skilled nursing facility, you get more ER visits.  
  • Who are the primary care doctors?  All PCPs are not created equal and defining the PCP base would require us to know what their baseline information looks like.  What is their payer source.  How many patients a day do they see?  Do they use extenders?  Do they leave open slots for hospital follow up?  Lots of unanswered questions about the practice style of PCPs who see their own patients in the hospital and the PCPs who defer their patients to the hospitalist.  These potential selection biases were not addressed. 
    • Are hospitalists admitting all the patients from the overflowing, underfunded federal sliding scale clinic with one doctor and 12 NPs and a schedule that takes four months to get any follow up appointment?  These patients  have a primary care physician on paper only.
    • Are the PCPs actually semi retired surgeons or cardiologists making a little extra money on their way out the medical establishment?
    • Are the PCPs who ask hospitalists to admit all their patients incapable of managing complicated medical conditions without extensive outsourcing to other subspecialists after discharge?
    • Are the PCPs referring patients to hospitalists too busy doing cash botox and other revenue generating procedures at the expense of good 'ol fashioned medicine.
    • Are the PCPs who refer to hospitalists doing so because they are too busy to take care of complicated medical cases?
    • Are the PCPs referring to hospitalists the kind that tell all their patients to go to the ED if they have any complaints because they can't squeeze them in to their already double booked clinic?
It's important, also, to put into perspective what kind of numbers we are dealing with when comparing hospitalists vs PCPs.  In the 30 day hospital discharge period
  • 18% of hospitalist cared for patients were readmitted, while 17.2% of PCP patients were readmitted
  • 6.6% of hospitalist patient had SNF charges while 6.3% of PCP patients did as well
  • 93% of hospitalist patients experienced any charge within 30 days of discharge compared with 93.6% of PCP cared for patients.
All our patients are sick and complex.  I think hospitalist patients are a touch sicker.  When you are dealing with medical problems at the extreme biological spectrum, it doesn't take much stress to decompensate into failure.  5% of our population spends 50% of our health care dollars. These folks are almost 80 years old.  We do stuff to these people we wouldn't do to our dogs.  Yet, many of these patients and families often expect immortality.  The ones that get it are the ones that say enough is enough.  Unfortunately, as doctors, we often fail to present them that option as well.  You'd be surprised at how many  are just waiting for our permission. 

I  think there are many confounding variables I see every day in my practice that explain why patients I care for cost more after discharge than a PCP cared for patients.  Even if I believed it, so what.  The hospitalist advantage is worth 57 million dollars a year to hospitals.  I'll take that advantage over the minor Medicare inconvenience.    We are valuable as  evidenced by the rapidly expanding support payment.  We are valuable to hospitals with the quality driven platform we provide.  In the end, it doesn't really matter whether hospitalists cost an extra $300 or not for Medicare. Because the cost of care is driven not by hospitalists, but by the illness our patients carry.  Since our government doesn't have the strength to stop the Medicare beneficiary all you can eat madness, we as physicians are simply going to have to stop offering patients therapies that have negligible therapeutic value.   If we want to save Medicare, we are going to have to start saying no to patients.

I say no every day.  That echo isn't indicated.  You don't need to see your pulmonologist because your lungs are doing just fine.  I know the nurse practitioner sent you from the small town ER to see a neurologist, cardiologist and gastroenterologist, but I'm sorry ma'am, it's not medically necessary.  You're goint to see me and probably go home the next day.  As the Medicare axe gets larger and larger, we are going to have to do less with less.  But how?  I have the perfect solution.  In every hospital in this country, every patient needs to be screened, on admission, with a palliative performance scale  If the patient fails to achieve evidence based markers of functional status, hospitals need to start declining access to both doctors and patients for certain  services, surgeries, procedures and testing as a way to deny access to unreasonable and irrational care.  And it needs to by systemized so doctors don't feel naked to an assault from the legal community brought on by angry patients and families.  If our government isn't going to say no, then we must. 

We live in a skewed land of fairies and fantasies, where a 2.4 trillion dollar deficit cut over ten years (a simple average of $240 billion dollars a year)  is called an economic  victory, when in fact, annual deficits are running one trillion dollars a year as far as the eye can see and the news is stuck on hospitalists costing $300 a year.  It's just insane.

Here's a few other takes on the issue

Print Friendly and PDF