Archive for May, 2013
Deaths from surgery increase from a low on Monday to a high on the weekend according to a BMJ study just released May 28, 2013. The researchers evaluated about 750,000 scheduled surgeries and suggest the likely explanation is poor quality care on Friday and weekends.
Friday surgery has a 44% increase in the odds of death within 30 days compared to Monday. Weekend surgery is worse with an 82% increase in risk.
The 3-in-1 hospital. All hospitals are divided into three hospitals because of nurse shift assignments, weekends, holidays and vacations. And, the tendency for senior staff to gravitate to the weekday shifts. The hospitals are:
- The day hospital
- Then night hospital
- The weekend and holiday hospital
The day hospital has a full staff with the most senior doctors and nurses available and the operating room working at optimal capacity.
The night hospital has a limited staff of doctors and nurses and limited operating room staff.
The weekend and holiday hospital — this problematic hospital is when the staff levels are low, the operating room is on a standby, the least senior nurses are working and the weekday doctors are replaced by on-call doctors usually doing double duty (often overworked, not familiar with the patients and sleepy from long night hours). Furthermore, on-call surgeons may try to delay treatment of surgical complications until weekday surgeons are available.
The real surprise in this study is that the death rate rises as the week progresses. The cause is not clear but some possibilities include fatigue, surgeons being less available to resolve hospital problems when at an outpatient clinic, afternoon golf, and possibly scheduling patients just discovered to have surgical problems that week in the ER (delay in care).
Another negative factor is the global surgical payment (surgeons receive one payment for a surgery including all pre- and post- surgical care). So, unless there is a very equitable system within a group the surgeons who round on weekends are not paid and have little incentive to do more than “howdy” rounds.
How to correct the problem:
- Hospital actions: spread the surgical service activity over more shifts and weekends. At least one third of nursing staff on any shift should have more than 5 years experience. Pay surgical hospitalists to attend to surgical patients when the primary surgeon is not in the hospital and to service the emergency room at any time.
- Surgeon actions: Provide the hospital with a surgical hospitalist who is at the hospital 24/7. Allow the surgical hospitalist to do surgery on patients admitted from the ER. Reduce the Friday and weekend workload with more physician staff on those days. Provide monetary compensation for covering doctors (not just payment in kind).
- Patient actions: Ask about whether there is an experienced surgical hospitalist available in the hospital at all times (not just a resident or medical hospitalist). If not, refuse elective surgery on Friday or on weekends.
Medical x-rays now account for about 15% of radiation exposure in the United States and the amount is growing. Intermountain Healthcare in Utah now has a program to keep track of the accumulated radiation given to patients with x-ray procedures.
So, what is the true story? We know the advances in CT scanners have transformed medicine — we can see internal organs that previously could only be viewed with surgery. Doctors can diagnose internal problems with much improved accuracy. But, we also know CT scans, like all x-ray tests, come with some radiation exposure. The true story is a large measure of good with a pinch of bad.
The bad aspect of radiation with medical imaging is the slight tendency to set off cancer. Not right away, but many years after the exposure. The slight increase in lifetime risk is much less than 1%. To keep it in perspective the lifetime risk of fatal cancer is about 25% (25 out of 100 people die of cancer). Clearly, other things besides x-rays are causing cancer. 15% of radiation exposure is from medical imaging, but 85% is from background radiation — natural radiation from our environment, the sun and the stars.
There are imaging methods that have no radiation. Those are MRI (very good and very expensive) and ultrasound (blurry and cheap). If MRI was fast and low cost it might do away with regular x-rays but research is a long way from that ideal.
At least for then next decade x-rays will be the work-horse for medical imaging. Below is a table listing the x-ray exposure for various x-ray procedures. A chest x-ray is one of the most common x-rays and fortunately one that takes very little x-ray radiation. So, other procedures are often compared to the chest x-ray, also seen in the last column of the table.
The milliSievert (mSv) is the measure of radiation dose. One chest x-ray gives about 0.34 mSv. That amount of radiation is what a person would normally accumulate in 52 days from background radiation alone. Radiation treatment intended to damage tissue (like cancer tissue) requires 20,000 to 60,000 mSv — medical imaging is not in that ballpark.
Review of CT scans suggests that 1/3 may have been unnecessary. For instance, a CT scan done because of a bump on the head without any symptoms except from the bump would be considered unnecessary. So, an unnecessary CT scan is unnecessary radiation. None of us want radiation without some benefit. This thought is in the spotlight now, especially in emergency rooms, where a lot of the unnecessary scanning happens.
It seems reasonable to keep a running total of radiation exposure from medical imaging. Use the above table add up the radiation from all imaging tests you have had — a good number to tell your doctor, especially when additional imaging is proposed. The question “is that imaging test worth the radiation exposure?” is certainly reasonable.
This premium and that deductible — which to choose? If you ever had to purchase health insurance it’s not all that clear — and you wonder, what are they hiding?
We all face the risk of serious illness and the associated high costs. The above graph is about the risk for a couple age 60 to 64. The horizontal axis is the possible health bill from zero to something very large. The vertical axis is the number of people who accumulated the health costs enumerated on the horizontal axis. Although the most likely bill is about $10,000 there is a possibility of any amount. Unless a person has an extra million dollars lying around we purchase insurance to cover the unexpected health bill (or go bankrupt).
An insurance company keeps detailed records of health care bills for people of of various ages. Point C on the graph is the mean (average). Insurance companies have thousands if not millions of customers. So, they don’t need to know what might happen to any one person, all they need to know is the mean cost for the age group. 100,000 people times the mean of $18,000 is $1.8 billion — that’s how much they will pay out in a year (plus or minus a small percent).
The insurance company knows that point C on the graph for the benefits they offer. Then they devise an insurance plan to collect that amount so they can pay customer health bills. You might think that amount of money is a secret but not really. Below is a graph of the offerings on the internet of a real insurance company in May of 2013 for 2 people age 60-64. The yearly premium is on the vertical and the deductible is on the horizontal. The blue dots are the plans offered.
Interestingly, the blue dots are on a straight line! The line B to C is that straight line through the offered products. Point A (numerically the same as point C) is that amount you would pay in premiums if there was no deductible — in other words “first dollar coverage”. Point C is the deductible you would pay to have no premiums (true, no insurance offers such a plan). Thus, point C is that mean cost per customer as in the top graph (it also includes the operational costs and profit for the company — usually about 20%).
What is Point B? — it lines up with the blue dots of the actual plan. It looks like the amount you would pay if the deductible was zero. Not so fast, the difference between Point A and Point B is the amount the customer is likely to pay in co-pays and co-insurance.
The gap between the red line and the blue line is paid by the customer if they use health services. The part the customer pays is usually limited by the “out of pocket maximum”. If the customer does things to stay healthy it gives them a chance to actually pay less.
Comparison shopping is very difficult. The plans and benefits from insurance companies are different. The best advice is to purchase from a top rated company with good customer satisfaction. If you don’t have much savings a low deductible plan has advantages. If you can stand occasional high bills a high deductible plan may work. Check out insurance company ratings by Consumer Reports and JD Power.
How do the Brits do it? They made a healthcare system with twice the quality at half the price compared to the US (according to the Commonwealth Fund cost per person per year US $7960 UK $3487, developed country quality rank UK #2 US #7). Simply, they do it by having original ideas and a willingness to adopt good ideas from other countries.
The National Health Service (NHS) of the UK was born in the aftermath of WWII. Taxes pay for the system, which is free to citizens at the point of care. Internally, the system is based on capitation — doctors and hospitals are paid by the size of the population they serve. The system grew to be one of the highest quality and least expensive systems in the world. In the 90’s it was bogged down by waiting lines and old facilities until a modernization push got it back on track around 2000.
The DRG example: In 1983 Medicare adopted a way to pay hospitals with a single payment for each case based on the diagnosis of the patient. This revolutionary idea was called the diagnosis related group or DRG. NHS experts embraced the Medicare cost saving idea and renamed it the HRG (health resource group) and started using it in about 2003. Consequently, by adopting what works, the UK has strikingly transformed the financial workings of the NHS.
The NHS noticed cost variations between providers and solved the problem with “Best Practice Tariffs”. That means if the provider follows a well established guideline they get a full payment, if not, the payment is lower. In the US we call that concept “value based purchasing” (VBP) but the US only has a few pilot projects and only dreams about making VBP happen on a large scale.
The UK decided they wanted better results. The reform was called Payment by Results (PbR) and implemented in 2013. The results they expected were high quality, adequate volume of services, and cost efficiency. The NHS basically tweaked the capitation formula with incentives for the desired results.
The US Affordable Care Act (Obama Care) encourages the aggregation of doctors and hospitals in an economic model called an Accountable Care Organization (ACO). The US thinks it invented the idea behind the “Accountable Care Organization”. Actually, the concept is just a spin-off from the Primary Care Trusts and Hospital Trusts in the UK which have been functioning for over 60 years. Think: “Trust”=”ACO”.
The recent “Perspective” in the New England Journal of Medicine (NEJM 668;16 April 18, 2013, page 1465-1468) describes the recent IOM report requested by Congress. The authors lament the “Geographic Variation in Medicare Services“. The NHS is well known for controlling health care costs. Looking across the pond to the UK, here are some references that might be helpful to them:
- A simple guide to Payment by Results
- A person based formula for allocating commissioning funds to general practices in England: development of a statistical model
- Payment by Results: time for a rethink?
- Regional variation in the productivity of the English National Health Service.
Some understanding about how the NHS works would also be helpful. The following diagram is an overview of how the NHS controls cost associated with hospitals and doctors. They also have a good system for dealing with drugs and devices — a good topic for a future blog.
(figure revised 7/11/13)
Other charts of organization can be found at Nuffield Trust – New Structure of the NHS slideshow, NHS website – new structure, and History of the NHS.
In the UK 90% of health care is controlled by the government and 10% by the private sector. The UK Parliament sets a budget for health care which is administered by the Department of Health. Based on the funds allocated in the budget the Department of Health makes a national price list for services (unlike the US where there is no cap on expenditure) .
The “SUS” approves payments to providers based on the national price list (national tariff) and adds the features of PbR (payment by results). The commissioners are the paymasters and transfer funds to the providers. The providers keep track of actual costs and must provide cost data to the Department of Health (unlike the US where real costs are proprietary information and hospitals use the infamous “chargemaster“).
The US could learn a lot from other countries. The NHS in the UK seems very willing to share what they have learned over the years — and it is in English. The old saying “America and England are two countries separated by a common language” is especially true for health care.
Two days ago the Centers for Medicare and Medicaid Services (CMS) published the fees hospitals charge uninsured patients. These fees are called the “chargemaster” of the hospital. Hospitals point out that most people have some type of coverage. That is true. But, 50 million people do not have insurance and the fees apply to them. The following are the statistics for Colorado hospitals for which CMS presented data on the charges for heart attack. The column in pink is the amount charged and the MC Allowed column is what Medicare has calculated as a fair payment.
Hospitals are businesses and can charge what they wish for services. Except, many hospitals are “not for profit“; which means the money they would normally need to pay in taxes is instead used to benefit the community. They often prove that benefit by writing off the money they can not collect from people who lack insurance.
The unfortunate effect is that hospital charges are the number one cause of bankruptcy. Furthermore, the extreme penalty of the “chargemaster” for people without insurance drives people to purchase insurance — certainly a happy situation for insurance companies.
If you go to any health insurance web site you can see what the “chargemaster” penalty actually is. Look at the charge for the highest deductible policy. That charge, in large part, is the fee to benefit from insurance company negotiations with hospitals. Of course, the hospital makes the charge very high so any mark-off for insurance companies is just for show. This is an old retail game — raise the price 100% then have a 50% off sale.
A simple solution:
If the charges by hospitals were restricted by law not to exceed 20% above the fee allowed by Medicare the cost of insurance would be substantially less. And, people who select a high deductible plan could really enjoy low rates with protection from the cost of catastrophic illness.
It is important to note insurance reform will NOT solve the US problems with high cost and low quality health care. Insurance just distributes financial pain over a greater number of people. The solution is to fix the underlying problem not distribute the problem to more people.
Denver Colorado is a metropolitan area with several hospitals so the city serves as good example of the dilemma people face to make a choice.
The Denver Metro area is lucky to have good hospitals. But, the experience one might have as an inpatient depends on many factors including the individual doctors, the nurses and the strength of the quality improvement program at each hospital. The Center for Medicare and Medicaid Services (CMS) sets quality targets for participating hospitals and measures how those hospitals perform. CMS publishes the data and several organizations extract the data and make them available online. One such site is WhyNotTheBest.org by the Commonwealth Fund.
Hospitals tend to emphasize heart care because it is profitable and it has a strong emotional appeal. However, the average consumer never really knows what will land them in the hospital. CMS provides many different hospital quality statistics but the “overall” statistic is actually the most helpful to consumers.
Quality can be measured by an “outcome” such as mortality. Or, quality can be measured by adherence to a “process” . Process measures are popular because they don’t depend on patient factors such as age or poverty.
For example, leaving home in the morning should include the process of putting on shoes and closing the door. Following that process does not guarantee a good outcome but lessens the possibility of humiliation at work or later finding a burglar in your home. Hospital process measures reflect the quality of management and the ability of an organization to execute a plan with numerous players.
The following table is CMS data organized by WhyNotTheBest.org. The “Overall” is a weighted average of all of the process-of-care, or “core” measures.
NAME — Selected Denver Hospitals
|University Of Colorado Hospital Anschutz Inpatient (CO)||99.38%||162 of 1792|
|Swedish Medical Center (CO)||99.17%||265 of 1792|
|Presbyterian/st Luke’s Medical Center (CO)||98.76%||460 of 1792|
|Exempla Lutheran Medical Center (CO)||98.59%||551 of 1792|
|Exempla Saint Joseph Hospital (CO)||98.33%||698 of 1792|
|Denver Health Medical Center (CO)||97.56%||1093 of 1792|
|Centura Health-st Anthony Central Hospital (CO)||95.62%||1619 of 1792|
Hospitals love the percentage statistic and hate the national ranking. There are hospitals in other states that indeed do score 100%. Just like climbing a Colorado mountain, the last mile is often the hardest.
99.38% or 95.62% means the hospital failed to follow a process 6 or 44 times out of 1000. One time out of a million would be better. Which hospital you choose may depend on where your doctor or insurance company sends you. But, if you have a choice the information from CMS may be helpful.