Medical Economics

The Starling Curve Of Physician Productivity

As physicians, we learn about the Starling curve in the first year of medical school. It expresses the relationship between left ventricular end-diastolic volume and cardiac output. The same can be applied to the relationship between physician RVU productivity and departmental/institutional health. The physiologic relationship dictates that as end-diastolic volume increases, cardiac output also increases, but only up to a point. At some level of high end-diastolic volume, the heart gets over distended and over stressed – resulting in a progressive fall in cardiac output. It is the so-called “going over the top of the Starling curve”. This is the mechanism of heart failure – you over-fill the left ventricle and the heart “bags-out” with a reduction in cardiac output.

Last week, our medical center had its annual review of the departments. Each department gets 1 hour to update the Dean of the College of Medicine, the CEO & CFO of the medical center, and various medical center leaders about their department. Big departments, such as internal medicine and surgery, get longer than an hour. During each session, the department chairman provides an update on physician productivity, faculty attrition, faculty gender percentages, annual procedures, inpatient & outpatient visit numbers, etc. It culminates with the department’s assessment of strengths, weakness, opportunities, and threats. Medical center leaders use this information for strategic planning for the next fiscal year budget and resource allocation planning. As the medical director of one of our hospitals, I attended the sessions which went all day for 3 consecutive days.

As the department chairs successively presented every department’s data, I was struck by another curve that has the exact same appearance of the Starling curve.

As background, our institution has a goal that physicians will produce at the 75th percentile of benchmark productivity. We use data from Vizient (formerly, the University Healthsystem Consortium, or UHC). The Vizient database collects information from nearly all academic medical centers on things like physician salaries and productivity (measured by work relative value units – aka, wRVUs). A physician who produces at the 50th percentile of wRVU productivity would be average among peer academic physicians in that particular specialty. A physician who generates annual wRVUs at the 25th percentile produces annual wRVUs at the level of the bottom 25% of other academic physicians in that specialty and a physician at the 75th percentile would be producing more annual RVUs than 75% of all academic physicians in that specialty. At the Ohio State University, we are held to a 75th percentile goal.

On the surface, that sounds awful, like we are flogging our doctors to produce more than doctors at other academic medical centers. But under the surface, this turns out to not be true. You see, the Vizient wRVU benchmark is based on what a 100% full-time clinician would produce. Few, if any, of our doctors are 100% clinical FTEs (Full Time Equivalents). Most of the medical faculty have some portion of their time reserved for academic pursuits – generally 10-20% of their time. This time is meant to be used for teaching medical students & residents, writing papers, preparing lectures, writing research grants, etc. The problem is that there is no money coming in from anywhere to cover that 10-20% uncompensated academic time. So, we have to fund that academic time out of our clinal income from taking care of patients.

Lets take an example of a physician who works 50 hours per week. This physician is 80% clinical time (40 hours a week) and 20% unfunded academic time (10 hours per week). If this physician produces wRVUs at the 75th percentile during the 40 hours per week that they are doing clinical care, it turns out that this is usually pretty close to a physician producing at the 50th percentile during a 50 hour clinical care week week. In other words, the productivity all pretty much balances out given that our physicians rarely are 100% clinical FTEs.

So, getting back to the annual departmental reviews, for most of the departments, about half of the physicians were producing at the 75th percentile. This seems appropriate – younger physicians need a ramp-up period before they get optimally efficient and productive, consultants and surgeons need a few years to build up a referral base, and inevitably even some seasoned physicians are going to be more productive than others. Ideally, you want the department to be producing at the 75th percentile, in order to allow for these variances in individual physician productivity.

But one of the departments had a striking graph. In 2015, 96% of the physicians in that department were exceeding the 75th percentile of wRVU productivity compared to other academic physicians in that specialty. In 2016, about a third of the physicians in that department resigned – mostly to go into private practice. I got to thinking that wRVU productivity is a lot like end-diastolic volume. And department/institution health is a lot like cardiac output. Increasing physician productivity is healthy for the department up to a point. But each department has a peak to its productivity curve, and if productivity gets too high, then the department is stressed, moral declines, and physicians leave.

It turns out that when it comes to physician productivity, more is not always better. If it reaches a point where academic physicians have to see more patients than physicians in private practice (but still get paid less than physicians in private practice), at some point, those physicians leave in order to work less and make more.

I don’t know exactly where the peak of the physician productivity Starling curve is. I suspect that it likely will vary from specialty to specialty depending on a number of variables. But my best estimate is that when a department’s average annual productivity exceeds the 85th to 90th percentile of wRVU productivity, bad things happen: physician burn-out, physician salary envy, lack of time to teach, etc.

Some physicians are going to always be extreme producers and aren’t happy unless they are in the 95th percentile of productivity. For those physicians, you get them a really nice office, publicly acknowledge them, and make sure that they get plenty of administrative/nursing support. Those are the physicians that give others in your institution the extra time that they need to teach and do research. But for most physicians, if you see their productivity getting that high, it is a sign that you need more physicians if you are going to retain them in the future.

February 8, 2017

Inpatient Practice Medical Economics

Is A Nurse Practitioner Cost-Effective?

One of the most common requests that I get this time of year is for a doctor or a service to ask for hospital support for a nurse practitioner or physician assistant. In each instance, you have to do an analysis to determine if adding an NP or PA for inpatient management is financially worth it. In Ohio, although PAs and NPs have different training, they have similar scopes of practice and are often used interchangeably. So in this post, when I refer to nurse practitioners, it can also mean physician assistants.

In most situations, you want to ensure that by using an NP, that you are at least breaking even with the cost of the NP by the revenue generated by the NP. In Ohio, NPs can write prescriptions and bill independently. In the hospital, there are two ways that NPs can provide care with regular daily visits. (1) They can do a “shared visit” so that they do part of the encounter and documentation and the physician does part of the encounter and also does part of the documentation – in this case, the reimbursement is 100% of the physician’s reimbursement. (2) They can do an independent visit in which case the physician does not need to see the patient or document anything – these are reimbursed at 85% of the physician’s reimbursement.

NPs can either be hired by the hospital or hired by the physician. The key difference is that if they are hired by the hospital, then they are a hospital employee and as such, none of their documentation can be used for the physician’s note in order to bill a daily hospital visit. Therefore, if a physician wants an NP to help with daily rounds and note writing, then the physician has to hire the NP. Otherwise, it is a Stark violation. This primarily applies to medical admission patients – since surgeons get paid by a global fee for a given surgery, they are not required to have the same degree of individual documentation for billing daily encounters and so documentation by a hospital-employed NP doesn’t affect the physician’s reimbursement for the surgery. For hospital-employed NPs, there are creative ways that the physician can lease a portion of the NP’s time from the hospital but the NP would still need to generate enough income to pay for the portion of time that the physician leases.

In our hospital, there is a bylaw that requires that a patient has to be seen by a physician daily. Therefore, having an NP on an admitting service limits them to doing shared visits – they cannot see patients independently without a physician also seeing that patient. However, on a consult service, the NP can see a patient without the physician also seeing the patients, since the admitting physician is also seeing/documenting a daily visit on that patient. Therefore, an NP on a consult service can either do shared visits or independent visits.

Lets take the situation when the NP is on a consult service and is seeing patients independently. The average NP salary is $100,000; add in 25% benefits and that comes to $125,000. NPs tend to usually work closer to a 40-hour work week so let’s say they see inpatients Monday through Friday and the physician covering the weekend sees all of the patients the NP was following during the weekdays. We will further assume that the NP works 46 weeks a year (4 weeks vacation and 2 weeks of holidays over the course of the year).

Therefore, the salary/benefits cost of the NP is $2,717 per worked week or $544 per worked day. In order to break even on the cost of that NP, the NP would need to generate $544 of revenue per day after expenses. In a private practice, there are relatively fewer overhead expenses but in an academic practice, there are a bunch of expenses, for example: Dean’s tax, departmental expenses, divisional expenses, malpractice, billing/administrative expenses, etc. All told these typically run about 21%. So, taking into account overhead, the NP would need to bring in $688 per worked day to fully break even.

In most practices, the physician will see the initial consult on a patient and the NP will see the return visits to that patient – a consult is usually a request for the learned opinion of an experienced specialist who has spent additional years of training to become an expert in an area of medicine and so the physician usually does the initial visit and lays out an impression and plan for that admission. So, we’ll assume that the NP is seeing only return visits and bills, on average, level 2 returns (CPT 99232) – Medicare pays $71 for this level of visits; adjusting this for the 85% reimbursement received by NPs for independent visits, this equates to $60 per encounter (a little less for Medicaid and a little more for commercial insurance). Therefore, based on Medicare reimbursement, the NP would need to see 11.5 inpatient return visits per day in order to pay for his/her salary. That would work out to about 40 minutes per return visit encounter which is very achievable (assuming that the consult service is large enough to support this volume of return visits).

Physicians have higher salaries than NPs and thus the cost per hour of a physician’s time is greater than the cost per hour of an NP’s time. Therefore, NPs can be cost effective when doing very time-intensive activities such as palliative medicine, smoking cessation counseling, diabetic education, etc. Also, you have to take into account what the physician will be doing if they don’t see the return consult visits. If the gastroenterologist will be able to do more colonoscopies or the cardiologist will be able to read more stress tests, then you can afford to lose money on an NP’s salary and still come out ahead because you are able to do a lot more of a more highly reimbursed activity than you otherwise would.

So, putting all of this together, what can we conclude:

  1. NPs need to see an average of 11.5 return visits per workday in order to break even financially.
  2. It can be cost-effective for an NP to see fewer than 11.5 return visits per day on procedure-oriented services such as surgery, cardiology, or gastroenterology since the NP frees up the physician to do more procedures that pay more per hour than return hospital visits.
  3. The practice’s payor mix affects the number of visits necessary to pay the NP’s salary – a practice with little Medicaid and a lot of commercial insurance may only need the NP to see 9-10 visits per day whereas a practice with a lot of Medicaid may need the NP to see 13-14 visits per day.
  4. It is financially more advantageous to have NPs do time-intensive activities (such as counseling, arranging follow-up testing, etc.) instead of having physicians do these.
  5. It is financially more advantageous to have NPs see uninsured/charity care patients since the cost of the NP’s time is less than the cost of the physician’s time.

If the NP is doing a shared visit (either with the admitting service physician or a consult physician), then the number of return visits needed to cover the NP’s salary is less – 9.7 per day. However, since the physician still needs to see each of these patients and do a component of the progress note documentation for each of these patients, that physician’s time now needs to be considered since all of the revenue from those 9.7 encounters will be going to cover the NP’s salary.

February 5, 2017

Medical Economics Medical Education

The Hidden Time Cost Of Being A Doctor

It takes a lot of time to become a doctor. And once you become a doctor it takes a lot of time to keep being a doctor. The amount of regulatory requirements per year are staggering. These add up to time costs and every doctor has to pay these time costs, regardless of the number of patients that you see. As you will see in this post, these costs add up quickly and result in those doctors who do a lot of teaching, research, or administration spending a disproportionate amount of their time meeting these requirements.

Protected time (for research, administration, or teaching) is highly sought and highly prized in academic medicine. It has to be – the only way to get promoted and get a salary increase is to do something other than clinical care of patients. There is the obvious cost of these activities: they don’t pay very well so if a physician is going to make anywhere close to a full-time clinician’s salary, then someone else has to contribute money. But there are hidden costs – those that no one ever talks about but that can eat away at your physicians’ productivity and suck the life out of an academic department.

They’re the fixed time costs that we all pay in order to do our regular jobs. Whether you are a 100% clinical FTE (i.e., a physician who only takes care of patients) or a 25% clinical FTE (i.e., someone who only spends 1 out of 4 working hours taking care of patients), you have to do these regular activities in order to maintain licensure and medical staff privileges. And they can add up… a lot. Let’s take a look at some of the more common of these:

  1. Continuing medical education. In Ohio, we have to do 50 hours per year of CME to maintain our medical license.
  2. ACLS (Advanced Cardiac Life Support). Required for many specialties; for others, ATLS (Advanced Trauma Life Support), or PALS (Pediatric Advanced Life Support) may be required. Preparation and classwork is about 10 hours every 2 years.
  3. CITI (Collaborative Institutional Training initiative). This is required for any physician who is involved in human subject research. Because this includes enrollment in trials and not just being a funded researcher, many/most academic physicians have to keep their CITI certificate up to date just to be able to assist clinical researchers by referring patients into clinical trials. It takes about 12 hours to do the program and it has to be renewed every 3 years.
  4. Department faculty meetings. At our University, these are mandatory and held quarterly – 4 hours per year.
  5. Division faculty meetings. In our division, these are mandatory and held monthly – 12 hours per year.
  6. Electronic medical record training. Initially, this takes about 10-20 hours. However, once you are facile with it, you just need to get trained on the periodic software updates – about 1 hour per year.
  7. Compliance training. At our hospital, we have a mandatory 2 hours per year for billing/coding compliance training to ensure that we are documenting our services and billing correctly.
  8. Hospital training. At Ohio State, these fall under “CBL” (Computer Based Learning) modules. These cover everything from what to do in a fire, to how to read a hazardous materials label, to the hospital’s sexual harassment policy. They vary from year to year but typically, it is about 10 hours per year.
  9. Hospital committees. I attend an enormous number of committee meetings but I get paid to attend them as a medical director. However, no one fully escapes committees and most physicians find themselves on a couple. I’ll estimate 15 hours a year.
  10. “Justify your existence forms”. These are part of the annual review that every academic physician has to fill out to document their annual clinical/research/publication/teaching/administrative productivity and describe how they have spent all of their time over the past year. Included in this category is the “promotion and tenure dossier” that all academic physicians have to complete periodically as they move toward promotion to associate professor to full professor. In our institution, if a physician is in the so-called clinical track, even full professors have to fill these out every 2-5 years in order to have their university contracts renewed. If you include the required face-to-face meeting with the division director or department chairman, the process requires about 6 hours per year.
  11. Emails. I get 50-100 a day – most physicians don’t get quite this many. Many of these are mass emails to all physicians. Some are worthy of reading (like weekly hospital news briefs) but a lot are garbage (like people who hit the “respond to all” button on every congratulatory email sent by a chairman to recognize a notable achievement by one of the faculty members). You have to at least open all of them and skim the first few sentences to see if you need to read the rest or if you can just click the delete button on your email program. Probably about 50 hours per year on average.
  12. Licensure forms. Medical license, DEA license, etc. Plan on 1 hour a year on average to fill these out.
  13. Surveys. We get surveyed constantly – from the College, from the hospital, from the department, from outside agencies. Most physicians don’t answer most of them because there are just too many. But some are inescapable – figure 2 hours per year.
  14. Board certification maintenance of certification. This includes required “MOC modules” that some boards require physicians to do every year and also includes the renewal board examination test (every 8-10 years depending on the specific board) as well as studying in order to pass the board exam. Although some of these activities can double for continuing medical education requirements, some can’t so figure an overall average is about 5 hours per year that can’t be included in CME.
  15. Employee health. This includes the time it takes to get your annual flu shot and the time it takes to do the annual infection control learning module, among other employee health & epidemiology requirements. Overall, 2 hours per year.

So, add all of this up and you get approximately 169 hours per year that every physician has to spend doing required activities just to be able to see a single patient or to see a thousand patients. Given that most physicians work about 56 hours per week, this equates to 3 weeks of time over the course of a year. Let’s assume a physician works 48 weeks a year (off 3 weeks for vacation and 1 week for the sum of all holidays for a year). A 100% clinical FTE would need to spend 3 weeks doing all of their required activities resulting in 45 weeks of patient care per year. A 25% clinical FTE (for example, someone who spends 75% of their time doing research or administration) would have 36 weeks per year doing research/administration leaving 12 weeks per year left over to do clinical activities. However, because that physician would need to spend 3 weeks of time on all of the above activities, they would only really be seeing patients for 9 weeks per year.

The reality is that most of us end up doing most of these activities during the evening or on weekends. But they still represent a huge fixed time cost to any academic physician. As a result, you can potentially get more clinical work from one 100% clinical FTE than you do from four 25% clinical FTEs.

February 1, 2017

Medical Economics Physician Finances

Do Happy Doctors Make Less Money?

I was reading over the 2016 Medscape Physician Compensation Report and was struck by some of the results. Every year, Medscape does a survey of physicians about their income, job satisfaction, demographics, etc. Last year, 19,200 physicians, responded to the survey and it unveiled some curious results.

Perhaps not surprisingly, the 6 specialties with the highest incomes were:

  1. Orthopedics ($443,000)
  2. Cardiology ($410,000)
  3. Dermatology ($381,000)
  4. Gastroenterology ($380,000)
  5. Radiology ($375,000)
  6. Urology ($367,000)

Equally unsurprisingly, the 6 specialists with the lowest incomes were:

  1. Pediatrics ($204,000)
  2. Endocrinology ($206,000)
  3. Family medicine ($207,000)
  4. Infectious disease ($215,000)
  5. Allergy ($222,000)
  6. Internal medicine ($222,000)

The real surprise came in the responses to the question “Would you choose to go into medicine again?”. The physicians in specialties that were most likely to respond that if they could do it all over again, they’d still go into medicine as a career were:

  1. Family medicine (73%)
  2. Internal medicine (71%)
  3. Rheumatology (70%)
  4. Pulmonary (69%)
  5. Infectious disease (69%)
  6. Pediatrics (68%)

The the physicians in specialties that were least likely to go into medicine again if they had to do it all over again were:

  1. Plastic surgery (47%)
  2. Radiology (49%)
  3. Orthopedics (49%)
  4. Urology (51%)
  5. Dermatology (53%)
  6. General surgery (54%)

Notice anything striking? The physicians in the lowest paid specialties were most likely to choose a career in medicine if they were just starting out again whereas the physicians in the highest paid specialties were least likely to go into a career in medicine again.

I’ve been thinking about this and came up with a few possible explanations. First, it could be that the highest paid specialties are the most grueling and stressful leading to greater burn-out. Second, there could be career selection bias if medical students choose specialties based on projected income rather than what they are passionate about. Third, it could be that having more money makes you lament the fact that you don’t have very much time to enjoy it.

However, I’d like to think that there is a fourth explanation. The physicians who were most likely to say that they would choose medicine again were in specialties where there is temporal continuity of the doctor-patient relationship. By that I mean that those physicians tend to have patients that they take care of for years and even decades and develop long-standing bonds with those patients.

In my pulmonary practice, I have patients that I have managed their asthma for 25 years. Patients that I’ve seen regularly since placing airway stents 20 years ago. Patients who are the children of my patients from years past. These are people who when I look at the next day’s office schedule, I look forward to seeing them again. Over time, you become vested in a patient’s health, in their life, and in their family. It is one of the great satisfiers in medical practice.

In recent years, we have been under increased pressure to increase productivity. I’ve often been asked by business leaders to increase my new-to-return patient ratio by seeing more new patients and transferring return patients to advanced practice providers to see to see for follow-up visits. From a short-term revenue standpoint, this makes total sense, because those new patient visits pay better and have a lot of down-stream revenue to the health system. But in the long-term, it is the return patient visits that create the bonds that make physicians say, “If I had to do it all over again, I would”.

For doctors, money can’t buy you job satisfaction but maybe the doctor-patient relationship can.

January 14, 2017

Medical Economics

Medicaid Reduces Emergency Department Use

A couple of months ago, Ohio Senator Sherrod Brown visited our hospital to do a press announcement about legislation that he and Senator John McCain were sponsoring. After the formal event, I spent a few minutes speaking with Senator Brown. I was telling him about how we had seen a decrease in our emergency department visits ever since Ohio enacted Medicaid expansion. He challenged me by saying that the published data did not support the contention that Medicaid expansion reduces ER visits and in fact, some literature suggested just the opposite, that states that participated in Medicaid expansion actually saw an increase in ER visits. My explanation was that our ER was different in that we had an extremely high percentage of uninsured patients prior to Medicaid expansion, our hospital is in what was previously a “primary care desert”, and that since Medicaid expansion, we have greatly increased our primary care footprint in the neighborhood. Senator Brown remained skeptical but the next week, I got a handwritten note from him thanking me for my comments and telling me that it was a good conversation.

I’ll preface this post by letting you know that I’m politically ambidextrous – I’m not registered with either political party and vote for whoever I think is the best candidate for an office, regardless of party affiliation. Over the past 25 years, we have had unbelievably great Senators: John Glenn, Mike DeWine, George Voinovich, Sherrod Brown and Rob Portman. Prior to October, I had never met Senator Brown but I found him highly intelligent and very quick thinking – he won me over.

So, after our meeting, I started questioning my own observations. And then the recent Ohio Medicaid Group VIII Assessment: Report to the General Assembly was released and it validated our hospital’s recent experience.

Ohio’s governor, John Kasich, enacted Medicaid expansion in January 2014. Prior to that time, our hospital’s uninsured rate (for admitted patients) was about 12.5%. That means that we didn’t get paid for 1 out of every 8 patients admitted to our hospital. The uninsured rate in our ER was even worse. It was bad for us as doctors to not get paid for providing medical services, but it was even worse for the patients. Most of them could not afford the medications that were prescribed in the ER and most of them did not have access to primary care physicians for basic preventive care. So, for example, if a patient came into the ER with a severe asthma exacerbation and we stabilized them with nebulizer treatments and some IV steroids, they couldn’t afford the inhaler prescription that they got (average about $300/inhaler) and so they bounced back in a week and then became unemployable because of work absences.

After Medicaid expansion, our hospital’s admitted patient uninsured rate fell to 2.3%. That’s huge. And the percent of poor Ohioans without insurance dropped from 32% to 14%.

However, since 2014, we saw a funny thing in our monthly financial reports. Our emergency department visits started to drop. OSU purchased our hospital in 1999 and we had seen a steady and striking increase in ER visits from 1999 through 2013 – we had projected and budgeted for a continued increase so this was a bit alarming. But although our ER visits fell off, the number of emergency squad arrivals per day were continuing to increase and the number of hospital admissions through the ER continued to increase. This means that we were continuing to see more of the sicker patients but we were seeing fewer of the less sick patients, the so-called “treat and release” patients. My belief had been that we were getting more of these patients into Medicaid coverage so that they could fill their prescriptions and get non-emergent care at less expensive primary care physician locations that they now had access to.

The newly released Ohio Medicaid Group VIII Assessment report now shows that what we have observed is real – that Medicaid expansion has reduced ER visits. For adults between 19 and 44 years old, ER use dropped from 1,557 to 1,279 per 100,000 – an 18% reduction. For adults between 45 and 64 years old, ER use dropped from 1,349 to 877 per 100,000 – a 35% reduction.

43% of Ohioans in Medicaid expansion reported a decrease in unmet health needs and only 8.3% reported an increase in unmet health needs. 48% reported an improvement in their overall health and only 3.5% reported that their health had worsened. The previously unemployed Ohioans reported that it was easier to look for work now that they had Medicaid coverage and those that had a job said that Medicaid made it easier for them to keep their job.

State-wide, 702,000 people are enrolled in Group VIII Medicaid from the Medicaid expansion. In Franklin County, 18.5% of adults under age 65 are enrolled in Medicaid – that is nearly 1 out of every 5 adults. The counties along the Ohio River are even more striking with 25-30% of adults on Medicaid. 

So what does all of this mean? From the hospital’s perspective, Medicaid expansion is keeping patients out of our emergency department. And for most of these patients, that is a good thing. Emergency rooms are for patients with emergencies and not for patients with minor problems that don’t have any other access to healthcare.

Some would argue that Medicaid expansion is excessively expensive and that it needs to be abolished. I would argue that without Medicaid, these same patients would still be getting sick, it is just that they don’t get good treatment for their conditions so rather than being cured of their diseases, their diseases would just smolder – making them perpetually clog up our emergency departments and making them unemployable. When they get sick, our ERs, hospitals, and doctors still have to take care of them on moral, ethical, and legal grounds and so they continue to consume our country’s healthcare resources so we all end up paying for it with higher healthcare costs to everyone else. In the long run, the cost to our society is greater without Medicaid expansion.

January 10, 2017

Medical Economics

America’s Declining Life Expectancy?

I got on the Social Security Administration’s Life Expectancy Calculator and found out that I’m probably going to live until 83.1 years. That means that October 2041 is going to be a bad month for me. Two weeks ago, the Center for Disease Control reported that for the first time since 1993, the U.S. life expectancy decreased by 0.1 years to 78.8 years.

Since the report came out, there has been a lot of speculation about the causes: increased opioid abuse-related deaths, the effect of obesity, etc. But if you look at the data, the death rate is up for heart disease, respiratory disease, accidents, stroke, Alzheimer’s disease, diabetes, kidney disease, and suicide.

Americans already live shorter lives than inhabitants of other economically developed countries. The OECD (Organisation for Economic Cooperation and Development) reports on the life expectancy in 43 countries – the U.S. ranks 26th, just above Chile, the Czech Republic, and Turkey. But will the reduction in American life expectancy save Medicare?

This year, the Medicare Board of Trustees determined that the Medicare Hospital Insurance Trust Fund was going to run out of money in 2028 and after that time, federal Medicare tax revenues will only be able to cover 87% of Medicare costs. The reason is that there is a bolus of baby boomers retiring and going on Medicare, plus people were living longer (at least up until this year).

One of the problems is that when Medicare was created in 1965, it set the retirement age that Americans become eligible for Medicare as 65 years old. At that time, the life expectancy from birth was 66.8 years for men and 73.7 years for women. Now, men live 9.5 years longer and women live 7.5 years longer. But the Medicare eligibility age has not changed and has remained 65. If you just take those people who actually make it to age 65, then in 1965, the average American would live 14.8 years on Medicare and now they live 19.4 years on Medicare. That is an increase of 4.6 years since the inception of Medicare, a 31% increase in the demands on Medicare.

So to make Medicare hold out, there are only a few options:

  1. Increase the Medicare tax
  2. Increase the Medicare eligibility age
  3. Reduce Medicare benefits
  4. Reduce Medicare costs
  5. Reduce the life expectancy

If this year’s reduction in American life expectancy continues as a trend in the future, then it would take 46 years to reduce life expectancy back to 1965 levels and Medicare will run out of money in 12 years. So living shorter lives probably isn’t going to work.

We could reduce Medicare benefits and require seniors to pay more out of pocket for their health care but given the increasingly large voting block of people over 65, I doubt that any congressman supporting this measure will stay in congress very long.

Reducing Medicare costs by reducing health care costs sounds great in theory but given our lack of success in the past 30 years, I don’t think we can count on costs coming down in the near future.

So, that pretty much leaves increasing the Medicare eligibility age or increasing the Medicare payroll tax (currently at 2.9% of wages).

My preference would be to tie the Medicare eligibility age to the U.S. life expectancy so that the projected average number of years an American would have Medicare coverage would be something around 18 years. With current life expectancy, that would mean increasing the Medicare eligibility age from 65 to about age 67. The bad news would be that you’ll have to retire 2 years later than your grandfather did but the good news is that you are going to live 4.6 years longer than he did. By increasing the eligibility age from 65 to 67, there will be two additional years that Americans pay Medicare payroll taxes and two fewer years that they are consuming the Medicare Trust Fund.

On the surface, this sounds like making Americans work more years. However, according to the U.S. Census, in 1965. about 50% of American adults completed high school; in 2015, 88% of adults completed high school. In 1965, about 10% of American adults completed college; in 2015, 34% of adults completed college. Therefore, Americans are completing their education and entering the workforce at an older age than they did in 1965 so by extending the Medicare eligibility age by 2 years to age 67, we wouldn’t be asking the average American to work more years than their grandparents did, we’d really be asking them to work the same number of years as their grandparents did.

So, the bottom line is that Americans don’t live as long as residents of other countries and we’re not living as long as we used to. However, our declining life expectancy alone will not save Medicare. We’ll have to take other measures.

December 24, 2016

Hospital Finances Medical Economics

How Many Researchers Can You Really Afford?

Academic medical centers’ reputations are rarely built on the quality of clinical care or the quality of education. Reputations are built on the volume of research grants and publications. The academic medical center becomes famous by doing research about clinical care and publishing about education. Similarly, to be promoted as a faculty member at most colleges of medicine in the U.S., it is not enough to be a great clinician or great teacher, you have to do research and publish about medicine and teaching. In theory, devoting a lot of time to research and publication about clinical care and education will also make the institution a better place to provide clinical care and medical education. But in reality, the best researchers and journal article writers are not necessarily the best clinicians or teachers.

To be successful obtaining and implementing research grants, physicians have to have “protected time”. This is time that they are not assigned clinical duties and can devote to scientific investigation and writing in order to be competitive for research grants. The most prestigious grants for physicians are those from the National Institutes of Health and these grants provide money to conduct research projects and also provide money to pay for the physician’s “protected time”.

But we have a problem in American research. Grants from the National Institutes of Health don’t really cover the physician’s protected time. It all comes down to something called the NIH salary cap. This is the maximum salary that can be paid from an NIH grant. Currently, the salary cap is $185,100. To any normal human being, this seems like a lot of money, a whole lot of money. The problem is, that physicians doing clinical practice usually make more than this. In fact, according to the MGMA salary survey, the average salary for most specialties is higher than $185,100. This means that to be a researcher, you either have to accept a lower salary than a clinician in the same specialty or someone else needs to subsidize your salary.

Lets take an example of a group of physicians who have 50% protected time, meaning that they see patients half of the work week and do research funded by the NIH for the other half of the work week. In the table below, the salary is taken from the MGMA survey. The the cost of 50% protected time is listed as 50% salary. The NIH salary cap is for a 100% full-time researcher is $185,100, so half of that (to cover the 50% protected time) is listed as 50% NIH cap. For most specialties, the NIH salary cap will not fully cover the salary that the physician would receive if her or she was a full-time clinician; this difference is listed in the last column.


From this analysis, you can see that a hospital can afford to have 4 specialties do research without having to subsidize them: infectious disease, general internal medicine, nephrology, and general pediatrics. For any other specialty, the hospital has to come up with additional funds to make up the difference between the NIH salary cap and what that physician could make doing pure clinical practice. Neurosurgeons are the most financially challenging since they have the highest salaries: you can fund 19 endocrinologists to do research for the price of funding one neurosurgeon.

In reality, most researchers accept a lower income than full-time clinicians. Researchers don’t have to round on weekends, don’t get called in at night for emergencies, and don’t have malpractice suits filed against them. But there are limits and even the most scientifically curious physicians will find the allure of an extra $50,000 or $100,000 too much to keep them in research.

As a consequence of this, an academic medical center that wants to get the greatest return on research investment will seek a lot of researchers from endocrinology, physical medicine, infectious disease, and nephrology. Researchers who are neurosurgeons, orthopedic surgeons, and cardiologists are too expensive to have more than a small number of researchers.

One of my colleagues who is a cardiologist on his division’s finance committee once told me that that the worst news he can get is a mass congratulatory email from the division director telling all of the cardiologists that one of their peers just got an NIH grant. The unwritten implication of that grant was that the rest of the cardiologists were going to have to pony up to help subsidize the portion of the grant awardee’s salary not covered by the NIH salary cap. Now days, the clinical physicians usually can’t afford to pay this difference because it means that they have to take a pay cut in order to support their research colleagues. Therefore it comes down to the hospital to provide the salary difference subsidy.

So as a hospital medical director, how should we view this? We only have a limited amount of money to invest in researchers so we have to be prudent in how we spend it and who we spend it on. It is like investing. For high salary specialties, the hospital can only afford a small number of researchers and they have to have a high probability of research success – think of this as buying 1 expensive stock share in Apple. For lower salary specialities, the hospital can afford a larger number of researchers and can afford to take a chance on researchers with a less certain probability of research success – think of this as buying 1 inexpensive stock share in each of 10 start-up companies.

The holy grail of research funding is the endowed chair where the academic medical center can use money from donors to off-set the NIH salary cap difference. This is pretty easy at a well-endowed college of medicine like, for example, Harvard. But it is not so practical at a state-supported college of medicine (like Ohio State) with relatively meager endowment funds. For institutions with less endowments, you have to decide what the right ratio of clinicians:researchers is. That ratio will vary depending on the specialty and the percent protected (research) time that the researchers have. The goal is to have the right balance so that you have enough research to make the institution famous but not so much research that institution goes into debt.

December 3, 2016

Medical Economics

Dismantling Obamacare

chopping-blockAfter the elections of 2016, the Affordable Care Act (aka, Obamacare), will no longer exist as it is today. Many politicians have vowed to repeal it immediately. Repeal would be pretty complicated but dismantling would be considerably easier. So what’s on the chopping block?

First, the ability to keep children on their parent’s health insurance until age 26. When the Affordable Care Act was passed in 2010, the country was still in the depths of a recession and children straight out of high school or college had a difficult time finding jobs and if they did find one, those jobs had relatively meager benefits such as health insurance. So 6 years ago, this feature of the Affordable Care Act was widely praised. But in 2016, the economy is much better (despite what aspiring politicians running for office in the past 6 months have told us). unemployment-rateThe U.S. unemployment rate hit 10% in 2010, the highest it had been in more than 25 years and young adults between the age of 18 and 26 were the hardest affected so it made sense to create access to health insurance for them at that time. Last month, the unemployment rate was only 4.7% which is nearly the lowest that it has been for the past 45 years. Young adults are no longer living at home with their parents, they are living on their own with jobs that have richer benefits, including health insurance. So now, it is not as necessary to keep young adults on their parents health insurance plans. health-care-costs-per-capitaOn the other hand, this is a pretty low-cost part of the Affordable Care Act. Adults in this age range have relatively few health care costs on average – they don’t require much health screening, they haven’t lived long enough to get too many diseases, and they don’t take many medications.

Second, the pre-existing condition clause. This was also very popular and prevented insurance companies from denying a person health insurance if they had a pre-existing condition. On the surface, this sounds like a great idea but you can only have it if you also have an individual mandate.

Third, the individual mandate. This is one of the most contentious parts of the Affordable Care Act, namely, that all persons had to buy health insurance, whether they wanted to or not. This went against the core of Americans’ sense of self-determination and freedom from government interference with their choices. The problem is, that you can’t have the pre-existing condition clause without the individual mandate. So if you get rid of the individual mandate, then you have to get rid of the pre-existing condition clause. Otherwise, no one in their right mind would buy health insurance until they got sick – why buy it if you’re healthy and just throwing your money away? It would be like waiting to buy life insurance on your spouse until after he or she was dead. The problem is that if everyone waits to buy health insurance until they have an illness or an accident, then the cost of that health insurance would skyrocket and likely exceed the overall health care costs of people over 85 of $32,000 per year (in the graph above).

Fourth, Medicaid expansion. This part of the Affordable Care Act is almost as disliked as the individual mandate. Many states opted out of Medicaid expansion because of their state legislatures hatred of what was perceived as an expansion of entitlements. The good news is that with the economy improving so much since 2010 and the unemployment rate falling by more than 50%, there presumably won’t be as many people on Medicaid as there were in 2010. But the bad news is that in any society, there will always be people who are poor and like it or not, those people get sick. Before the Affordable Care Act, about 12% of patients in our hospital were uninsured, which basically meant that they couldn’t pay anything. But we as hospitals and physicians are required by law to take care of those patients which means the doctors’  time, the nurses’ time, the lab tests, the hospital food, the x-ray tests, the operating room time, and the medications all had to be provided for free. Can you imagine running a restaurant where you were required by law to feed 12% of your customers for free? Or maybe a car shop where you were required to provide free car parts and service for 1 out of every 8 people who come to your shop? You’d go out of business in 3 weeks. The reason that hospitals didn’t go out of business before the Affordable Care Act was the DSH (disproportionate share hospital) funds that provided hospitals Federal funds if they provided a lot of free care to the uninsured poor. The problem is that the DSH funds were reduced as Medicaid expansion took place as part of the Affordable Care Act and unless those funds are re-instated, then hospitals with a lot of uninsured/poor patients stand to go out of business (and then the uninsured poor will just go to another hospital that hasn’t gone out of business yet).

Fifth, the contraceptive mandate. Oh boy, did this one fuel anger. Religious organizations, that are historically on the side of more liberal social justice issues, were burning at the thought of having to pay for birth control pills when church doctrine opposes birth control. But here’s the thing: birth control pills are not terribly expensive, about $15 per month or $180 per year. On the other hand, the cost of a pregnancy (including pre-natal care, delivery, and post-natal care) is about $8,800, that is fifty times more expensive than a year’s worth of birth control (and that doesn’t even include the cost of maternity leave). Many of the same groups who don’t want to pay for their employees’ birth control pills also want to make abortion illegal. My own opinion is that, with the exception of rape, mother’s health, and minors, most abortions are a symptom of inadequate access to birth control and inadequate education about birth control. So if you really want to reduce abortions, then increase access to birth control. And if you really want to save money and reduce abortions, then make IUDs accessible – they cost about $800 all told and last for 12 years plus they are more effective than birth control pills. If you depreciate the cost over 12 years, it works out to about $65 a year. So, the reason to get rid of the contraceptive mandate would have to be a moral one because from an economic standpoint, it would make absolutely no sense. chance-of-pregnancy

Sixth, accountable care organizations. These were created as a part of the Affordable Care Act as a mechanism for hospitals and physicians to find low-cost ways to care for a large group of patients by promoting disease prevention, reducing unnecessary testing, and reducing unnecessary surgery. A lot of hospitals and physicians have been wary of ACOs and our health system, like many health systems, did not jump straight in head first to form an ACO. In the past 2 years, many ACOs failed to save money and as such the hospitals and physicians had reduced Medicare payments and basically lost money.

I have absolutely no idea about what is going to happen with heath care legislation over the next 4 years but we can be sure that something is going to happen. Currently, even with the Affordable Care Act, the United States has more uninsured citizens than just about every other industrialized country, we pay more for health care per person than any other country, and by most metrics, we have lower quality of health care than the majority of industrialized countries. Whatever we do in the next 4 years, I hope it is done with the goal to reduce our health costs and improve our health quality and not just to seek revenge on those that created and passed the Affordable Care Act.

November 13, 2016


Medical Economics

Penalizing Survival

There is one constant in regulations and policies… unintended consequences. Presently, Medicare seeks to improve quality of care delivered in hospitals by rewarding those hospitals that have lower mortality rates and lower 30-day readmissions. On the surface, this sounds like a great idea. But under the surface, are we penalizing hospitals that keep more patients alive?

In fiscal year 2013, Medicare began penalizing hospitals for higher than average 30-day readmission rates for patients with heart failure, myocardial infarction, and pneumonia under the Hospital Readmission Reduction Program. In fiscal year 2014, Medicare began penalizing hospitals for higher than average 30-day mortality rates for patients with these same diagnoses under the Hospital Value-Based Purchasing Program.

Here is the problem. Patients who get readmitted frequently are more likely to die in the near future – they are sicker. Conversely, patients who are sicker and at risk for death but are kept alive by better in-hospital care are more likely to be readmitted – they too are sicker. If Medicare equally penalized excessive readmissions and excessive mortality, then these factors would balance out. However, readmission penalties are 10 times greater than mortality penalties.

A recent study in JAMA Cardiology showed that this discrepancy in Medicare penalties does result in greater penalties for those hospitals that keep sicker patients alive than for those hospitals where equally sick patients die. In other words, Medicare disproportionately penalizes those hospitals that keep their mortality rates down at the cost of having higher readmission rates.

Last year, our hospital had an exceptionally low inpatient mortality rate of 0.54 which puts it at the second lowest mortality rate for all academic hospitals in the United States. We also struggle with a higher than average readmission rate. The implication of this study is that hospitals like ours keep patients alive but they live only to be readmitted another day. The results of the study further suggests that we would have gotten paid more by Medicare if more of our patients died so that they would not live to be readmitted.

The solution is to equalize the Medicare penalties for excessive readmission and for excessive mortality. I think that most patients would prefer to stay alive and be readmitted within 30 days instead of dying and not being readmitted.

November 5, 2016

Medical Economics

Kicking Hepatitis C Down The Road

Hepatitis C has exposed one of the larger cracks in American healthcare financing. In particular, the drug Harvoni (ledipasvir-sofosbuvir) has shown us the inherent conflict between private health insurance (commercial insurance companies) and public health insurance (Medicare, Medicaid, etc.).

Hepatitis C is an enormous problem in the United States. There are about 2.7 million Americans with chronic hepatitis C – thats 1% of our population. Worldwide, there are about 200 million people infected with the virus. It is the #1 cause of liver transplant in the United States and it causes around 10,000 deaths per year. Half of the people infected don’t know that they are infected and there is no vaccine to prevent it. The most recent cost estimate for the virus in the United States is $6.5 billion per year. That is $21 for every man, woman, and child in our country. In other words, this is a big public health problem and big cost to our nation.

But things are changing for hepatitis C. Patients can now be essentially cured of the infection with Harvoni. The problem is that Harvoni costs $90,000 for a 12-week course.

So, here is what happens. A person acquires hepatitis C as a relatively young adult (when they have commercial insurance) and then the hepatitis C manifests itself 20-40 years later with cirrhosis or hepatocellular carcinoma as an older adult (when they have Medicare).

The commercial insurance company is financially motivated to prevent conditions that would arise when a person is relatively young and still covered by that insurance company, because those conditions are costly to treat. Therefore, it is cheaper for the insurance company to prevent a disease that is going to show up in 3-4 years than to have to treat that disease 3-4 years later. A example of this is an insurance company paying to treat high cholesterol with a statin in a 40-year old so that they don’t have to pay for the person’s myocardial infarction when they are 44-years old.

For hepatitis C, the insurance companies have little financial motivation to treat since the company isn’t going to be paying for the liver transplant or hepatectomy when the disease finally manifests itself – Medicare will. The insurance company is incentivized to get you to 65 without any expensive medical problems. There is no incentive for anything that happens to a person after age 65 – that’s Medicare’s problem.

On the other hand, Medicaid is incentivized to treat hepatitis C. The prevalence of hepatitis C is 7.5 times higher in the Medicaid population than in the commercially insured population and persons with hepatitis C on Medicaid are much younger than those who are commercially insured. Medicaid’s goal is to keep the person healthy enough that they can eventually become gainfully employed and therefore get off of Medicaid and onto a commercial health insurance plan. If that person develops cirrhosis, they likely won’t be working and will stay on Medicaid until they die, racking up higher costs for Medicaid.

Somehow we need to change the incentives so that all insurers are motivated to keep people healthy not just to age 65 but beyond. This will not only ultimately reduce healthcare costs in the United States by reducing the costs incurred by older Americans covered by Medicare but it will also keep Americans healthy enough that they can continue to work after age 65 which will reduce poverty among older Americans and put less financial burden on our nation in the form of Social Security. When the commercial insurance company denies coverage for Harvoni, it is not because the insurance company doesn’t think Harvoni will improve your health, it is because by not covering it, it will improve the company’s financial health.

We as Americans like to think that we have the best healthcare in the world. And as a Cleveland Brown’s fan, I keep thinking that this year we’re going to have a winning season. Somehow, reality always catches up.

November 3, 2016