KOLAHUN

Too much health insurance in America...

I would like to update the SCHIP debate and the insidious restrictions Bush wants to place on them...but that'll have to wait until tomorrow.

In the meanwhile, feel free to read this wonderful National Review evaluation of the Republican presidential-hopefuls' health care plan and their desire to allow tax-incentives for personal health care purchasing as opposed to a) government sponsored and b) employer paid:

The centerpiece of Giuliani’s plan is the proposal, first made by President Bush in January, to change the tax treatment of health insurance. The tax code currently encourages people to get health insurance through their employers rather than to get higher wages and to pay for their insurance themselves. It provides a larger tax break the more expensive that insurance, encouraging people to choose policies that offer greater coverage. Bush, and now Giuliani, would extend the tax break to individuals who purchase their own health insurance. They would provide the same tax break regardless of the cost of the insurance, so that purchasers would have more incentive to shop around.

Romney proposes different changes to the tax code. He would make all out-of-pocket health expenses deductible against income taxes. That change would reduce one distortion of the economy: People would no longer have as great an incentive to get insurance to cover routine medical expenses rather than to pay for them out of pocket. But there are two problems with this proposal. First: Judging from the work of his principal health-care advisers, Romney’s plan would still give employer-provided plans an advantage, as they would be exempt from both income and payroll taxes. Giuliani’s plan would level the playing field more; individual ownership of health insurance would no longer be penalized at all.

Clearly then, the problem with the American health care system is TOO MUCH COVERAGE!  Gulliani believes that allowing employer based deductions incentivizes more expensive plans and Romney feels that well...people really don't need insurance coverage for routine medical expenses. 

Yes, I too hear the same complaints from small business owners - their employees are choosing too pricey a program...and from women that if only they could pay for their own Pap smears, our health care system would once again work.   

Un-complexity

I love the VA health care system for many reasons, not the least of which is it's simplicity.  The VA hires its own physicians, so all referrals are in house; it draws its own blood; it has its own pharmacy (with a few types of each class of medicine. 

Compare this with University Hospital in New Orleans:

• two medical schools provide residents and attendings for the hospital (LSU and Tulane) and patients are assigned a T or L number, meaning they can only get services by one or another school
• some services (e.g. orthopedics) see patients on alternate days, in lieu of T or L number
• some services have no counterpart in T or L (oral-maxillo facial surgery)
• the ICU has 2 L teams and only 1 T team
• some cardiology patients get assigned an L or T cardiologist based on acuity of illness rather than primary T or L number

If you're not fully sure what any of the previous bullet points means, imagine the frustrations of the house staff/ER staff who have to navigate this system.  Add the byzantine, incomplete outpatient clinics in this city and one can only imagine the inefficiencies of this system, the confusion it creates for the patients and providers and the overall disutility in its design as a health care system...

...oh to have the VA system back (at least for one self-sufficient hospital in this city!)

No, Dr. T...

Yesterday, Dr. T displayed quite a bitter antipathy towards using likelihood ratios.  He insisted on arguing that the goal medical tests should be to have the highest sensitivities and specificities.  Let's do a though experiment using chest pain to see which one is better.

(for a refresher on likelihood ratios see this  post)

I'll take the liberty of setting a couple of (nearly correct) assumptions for this post:

1. the threshold for performing an angiogram will be a probability of 70% that our patient has coronary artery disease causing the chest pain
2. the sensitivity and specificity of a stress test will be 60% and 90%, respectively

Here are our three patients:

• Patient 1 - 74 yo male with a history of CABG and 1 day of his typical, but worsening chest pain.
• Patient 2 - 48 yo male with h/o hypertension with progressive, squeezing, substernal chest pain
• Patient 3 - 24 yo college student with acute chest pain after failing her chemistry test

After taking a history and physical (and obviously ensuring that our patient does not meet WHO criteria for a heart attack using biomarkers and EKGs), let's estimate for each patient the probability that diseased coronary arteries caused the pain (i.e. pre-test probability):

• Patient 1 - ~90% (um...he's got a history of CAD with similar sx)
• Patient 2 - ~50% (it's hard to say - he's got risk factors)
• Patient 3 - ~1% (it's very likely she's having a panic attack rather than a heart attack)

Having said that, we can then use the sensitivities and specificities to calculate the Likelihood ratio (=sensitivity/(1-specificity)) of this particular stress test = 6.  Multiplying our pre-test probability for each patient by the likelihood ratio of the test reveals the post-test probability for each patient having coronary artery disease chest pain:

• Patient 1 - ~99%
• Patient 2 - ~85%
• Patient 3 - ~6%

(note that these are the probabilities if we had done a stress test on each patient and it came back positive.)

Now let's go through each patient:

• Patient 1 - he already had a pre-test probability greater than our 70% threshold we set, so we shouldn't do any further tests...he should go straight to an angiogram. 
• Patient 2 - now the stress test would be damn helpful for him.  He had a pre-test probability that did not cross our threshold.  But using a further test, if it came back positive, his probability would cross the 70% threshold and he would now get an angiogram.  In him, the test actually changed his medical management. 
• Patient 3 - even if she had a positive stress test, her probability (i.e. post-test probability) wouldn't cross our 70% threshold...so there's no reason to test her.  Even if it came back positive, we wouldn't do anything about it.

So now, to answer Dr. T, we can understand the clinical utility of likelihood ratios and assigning probabilities.  While he is correct that we desire higher sensitivities and specificities for our tests, these characteristics don't actually have any meaning to our patients.  For each patient above, the sensitivity and specificity were the exact same for each patient. 

But by converting them into a likelihood ratio, assigning pre-test probabilities, multiplying these two and setting a threshold - we can actually determine in each individual patient whether to use a particular test or not.  By just using sensitivities and specificities, we may understand the properties of the test - but miss how the test changes the medical perception of the patient.

Eyes wide shut

It's all nice and good that we've had a visceral anti-Michael Vick and anti-dog fighting response in this country, but it only brings about further questions about humanity's violence.  How many people reacting angrily toward Vick have cheered on Joyce Gracie tear a ligament or Randy Couture pummel his opponent in mixed-martial arts. 

...Or shoot defenseless deer or quail?

...Or bomb the hell out of each other for no apparent reason?

...Or perhaps even seen hulking, grown men  attack each other, drugging themselves up each weekend in the hope of feint glory and riches, while debilitating themselves in the process? 

Medical misnomers (volume 2)

Lupus anticoagulant - well, it's not usually associated with lupus, nor is it an 'anti-coagulant'.  In fact, it's a pro-coagulant...

Medical misnomers (volume 1)

Go-Lytely...not so much

Dog days of August

August in New Orleans sucks.  The brutal summer beats us down, hurricane warnings cause frequent evacuations (and sometimes brutal damage), and restauranters and musicians travel elsewhere until the fall.  But now, I fear that for the next few years we'll be subjected to a further insult: the political Katrina rememberences.

Barack Obama already showed up...President Bush will be here for the anniversary on the 29th and scores of other politicians (Hillary Clinton, John Edwards, etc.) will show their face.  And to what end?  Some New Orleans New Deal?  An influx of money, a plan to fix the levees, return people to their homes and the musicians to their gigs? 

Well, that's what local historian, Douglas Brinkley hopes in this all-encompassing Washington Post editorial today (it's worth the whole read):

Theodore Roosevelt set aside 230 million acres for wildlife conservation (plus built the Panama Canal). Franklin D. Roosevelt began a kaleidoscope of New Deal programs to calm the Great Depression and Truman oversaw the Marshall Plan rebuilding of Western Europe after World War II. Bush could seize the initiative and announce a real plan to rebuild, a partnership between the government, Fortune 500 companies and faith-based groups.

And yet, as much as I'd like this to happen, it saddens me to know that it just won't.  Look - let's leave alone Bush for a bit...the less said about him on this site the better.  Even Obama's plan to help rebuild this city does little more than shuffle around the chairs on the deck.  And the Fortune 500 companies?  The decades before Katrina saw a steady exodus of these companies out of New Orleans. 

But there's no reason to be nihilistic. There's a huge influx of latins workers and young intellectuals...and mixed with the amazing culture and creativity of the city itself - who knows what may emerge...

It'll be different, but will still be the most unique and precious city...

...and plus, in a few days August will end. 

Headline folly

The headline of this Times-Picayune article about health care in post-Katrina New Orleans misses much larger points about the health care system, the providers and the patients in this city:

"Without Charity Hospital, the poor and uninsured struggle to find health care"

Um...the poor and uninsured in New Orleans struggled to find health care even with Charity Hospital.   

At one point in the 20th century, Charity could boast about being a pre-eminent American Hospital.  But years of budget cuts, lack of upkeep and the city's overall financial decline decimated its quality well before Katrina.  Moreover, the pre-eminence of a single hospital as the sole provider of health care came into question even while Charity operated at its peak.  An ACP article summed up this dilemma in 1999 (6 years before Katrina):

"Charity Hospital is such a critical component of New Orleans' health care because it serves one of the nation's largest metropolitan concentrations of the poor and uninsured. Louisiana ranks fifth in the country in the percentage of its population without health insurance; its poverty rate is the fourth worst in the nation. In New Orleans, 22% of the residents have no insurance and just over half of the city's adults get health insurance through their work. (Many of the city's jobs are in the relatively low-paying tourism and service industries.)

Add the fact that there is no effective system for delivering primary care throughout the neighborhoods, and it is not surprising that the "underlying health status indicators here are terrible," said Kenneth Thorpe, PhD, of Tulane University's Institute for Health Services Research. A 1998 report published by the state identified Louisiana as 48th in the country based on various health indicators, including higher-than-expected rates of cancer, diabetes and infant mortality." (Emphasis mine.)

The public health officials in this city never worked out a way to provide primary care while Charity Hospital existed.  As a consequence, most patients ended up in Charity's waiting rooms, "bringing a breakfast and lunch", for end-stage disease treatment rather than preventive or early-stage treatment.  So while the headline of the Times-Picayune article tugs for tears, the body of the article actually contains hope:

"This is the health care safety net in post-Katrina New Orleans: patchy and provisional, propped up by miracle and chance, and heavily reliant on the charity of high-minded doctors, nurses and social workers who scrap for ailing patients, often by calling in favors with physicians or private hospitals. As with so many aspects of storm recovery, the hard work of ordinary citizens has substituted for an institutional solution."

While the public health officials dawdled after Katrina, in their vacuum a disparate group of health care workers built  the community clinics sorely needed (even before the storm).  While the early days of these clinics could be rightly described as "charity", these clinics - Common Ground, St. Thomas, the Tulane University Community Health Center at the Covenant House, St. Cecylia, the Woman's Clinic, et. al - have matured into more sustainable, professional clinics.  They have sought various funding sources, joined alliances and networks, began charging and taking insurance and have become an important cog in the health care system. 

So now, poor and uninsured patients can reasonably expect to be seen in a community clinic instead of waiting in the Charity Hospital ER.  In some senses, this marks a huge leap forward in primary care in New Orleans.  And while I see massive holes in specialty and tertiary care (patients have to travel an hour for chemotherapy, ENT, ophthalmology, orthopedics...), let's not sugar-coat the pre-Katrina Charity system. 

The Magic of New Orleans (volume 2)

About eight blocks from my place in the Treme lies one of the most amazing breakfast/lunch restaurants: Lil Dizzy's Cafe.

From catfish and grits, to all types of crawfish, shrimp omelettes, et. al Lil Dizzy's cooks up amazing food.  Little did I know, however, that the Baquet family belongs to New Orleans "royalty" as explained in this NY Times article. 

(Wayne Baquet can be seen everyday at Lil Dizzy's with his towel usually thrown over his right shoulder...but I now remember why his brother, Dean, seems familiar - he was featured on this PBS Frontline interview and video (on youtube) as former editor of the LA Times/current Washington Bureau Chief of the NY Times.)

Logisticians - my heroes

In my last week in Liberia, I half-jokingly asked to return to Medicine Sans Frontieres (MSF) as a logistician instead of a medic:

Logisticians provide support to MSF's medical programs. They co-ordinate the purchase and transportation of supplies, both locally and internationally; oversee transportation and communications equipment use and maintenance; and supervise local non-medical staff. They are also responsible for security protocols. Their duties include frequent interaction with local authorities and organizations. 

They're pretty bad-ass people...security, construction, transportation, etc.  And even in a medical NGO like MSF, they matter just as much as the medics - mainly, because if there are no meds or supplies...um, the medics can't do much.

Which brings us to one of the continued problems of emergency/relief care: logistical limitations.  Unless a disaster can be anticipated (e.g. hurricanes), it's nearly impossible to organize relief efforts within the first 48 hours - when a significant amount of the morbidity and mortality occur. 

For example, in the recent Peru earthquake which occurred on August 15th...MSF (an organization which has superb logistical capabilities) couldn't get the assessment team or the first batch of supplies onto the scene until August 18th:

August 16, 2007 – In the evening of Wednesday, August 15, the Peruvian coast was hit by a powerful earthquake (8.0-magnitude on the Richter scale). According to local sources, more than 500 people were killed and 1,000 were injured. The most affected cities are Chincha, Pisco, and Ica, located around 200 km south of the capital, Lima.

Doctors Without Borders/Médecins Sans Frontières (MSF) is organizing a team to assess the heavily hit areas in the south. A cargo charter is scheduled to leave Saturday, August 18, from Bogota, Colombia, with five tons of medical and non-medical supplies, including tents to set up dispensaries, and water-and-sanitation material.

Even with pre-packaged emergency supplies and ready-response teams, disasters which happen in remote locations cannot be accessed easily.  Relief workers need airports with large enough runways, either accessible roads + vehicles or teams of helicopters, an understanding of local conditions/languages/customs, local contacts AND enough security.  I daresay these requirements cannot be met within the first 24 hours...limiting the ability of relief workers to impact dire conditions in the early hours. 

Combined with the unlikely ability of poorer countries (or richer countries - see here) to have adequate domestic emergency responses, this situation, simply, blows for disaster victems.