Category Archives: Health Economics

Don’t miss the gorilla!

Posted on by

A recent narrative on improving value in health care (NEJM 2013; 368: 959-62) made use of a few psychological experiments to make some excellent points.

The first experiment asked the study participants to focus on a video and to count the number of passes for a team of three basketball players dressed in white while another three-player team dressed in black also played.  “During the video, a woman in a gorilla suit walked across the screen pounding her chest, remaining there for 5 seconds….Consistently, about 50% failed to notice the gorilla.”  Afterwards, when shown the video a second time, many of those who did not see the gorilla insisted that the video had been altered.

The second experiment involved perception.  “One group was initially shown a blurry image of a dog.  Then both groups were shown the exact same image, less blurry.  Those seeing the image for the first time had a much easier time recognizing it as a dog than those who had received previous ambiguous information.”  So, like those in the first experiment who did not  see the gorilla, participants are reluctant to change their opinion even when better information becomes available.

The author notes that the ‘value narrative’ can split patients and physicians into separate teams.  For physicians, value suggests avoiding overuse and providing evidence-based care.  For patients, value means enhancing their experience and catering outcomes that matter to them.  There is a study which shows that patients rate their care as better if they went to an ER for abdominal pain and underwent a CT scan (regardless of whether it was indicated).  Similarly, for back pain, two studies have shown that patients are far more satisfied with their care if they undergo an MRI.

The implications of these experiments on perception is that value in health care needs to be more transparent.

  • “We must admit that turning health care into a customer-service industry may to some extent undermine the delivery of evidence-based care.”
  • We actually know very little “about patients’ values and about how they should or might influence our decision-making.”
  • If we want to control costs and improve quality, “it will first require a look at the whole picture — and then a willingness to believe what we see.”

Off-label medications in pediatrics

Posted on by

In the U.S., multiple legislative changes have been improving the information available for off-label use of medications (JPGN 2013; 56: 113-4).

The Pediatric Research Equity Act (PL 108-155) and the Best Pharmaceuticals for Children Act (PL 107-109) help the Food and Drug Administration (FDA) promote pediatric trials and better information.

  • Pediatric data must be included in labeling whether results are positive, negative or inconclusive
  • If pediatric trials are not completed because the drug would be ineffective or unsafe, this information must go into labeling
  • Public internet posting of certain FDA pediatric reviews is required regardless of whether the trials led to an approval

Despite these improvements, the data may be difficult to interpret.  The authors of this commentary note that only “48% of trials of products that had pediatric safety information added to the product’s labeling were reported in the peer-reviewed literature.”  Furthermore, in about half of the published articles, the studies did not emphasize the same information as the FDA labeling.

Even when a fair amount of data is provided, it is frequently ignored.  A specific example given by the authors: lansoprazole. Lansoprazole is not approved in patients younger than 1 year and labeling describes a negative trial in this age group. Yet, in the U.S., lansoprazole was dispensed off-label in this population approximately 358,000 times in 2010.

Pharmaceutical companies benefit whether the prescription is for an approved indication or for an off-label use.  Even with approved uses, there have been substantial concerns about adverse effects.  The likelihood of more risk and less benefit is much greater for off-label use.

So with the next prescription, it might be worth thinking about whether the medication works, whether there is sufficient data to support an approved indication, and how much risk is involved.

Related blog posts:

Increased complexity or improper coding?

Posted on by

Not surprisingly, the adoption of electronic health records (EHR) has been associated with higher charges, especially in the emergency room setting (NEJM 2012; 367: 2465-67).  The question remains whether this increased complexity that is being billed is justified or simply due to “gaming” the system with electronic technology.

A specific example has been the increase in the highest level 5 codes used in emergency department (ED) visits, from 27% to 48% of Medicare charges (2001 to 2010).

While “gaming” may be part of the answer, more of the answer lies in the increasing complexity of patient conditions and more complex therapeutic options.  Specifically, the increase in higher coding has been associated with the following:

  • Marked increased use in new diagnostic technology. Overuse of technology is ill defined and failure to diagnose carries a much heavier penalty for physicians.
  • Increased numbers of patients without medical homes.  This increases the need for more diagnostic certainty.
  • Reduced hospital capacity.  Patients may be boarding in ED for quite some time before eventual admission or discharge.

Although the reasons stated explain how there may be more justification for higher complexity, it is well-recognized that EHRs facilitate billing by presenting check-boxes to more easily satisfy coding requirements.  This helps eliminate undercoding.  One of the drawbacks, though, with EHRs has been elaborate documentation at the expense of more direct patient contact.

Take home message:

The current system of coding is flawed.  While the complexity of care has increased, it remains difficult to identify the true value of the care provided. When most incentives reward increased testing and increased documentation which are easy to quantify, this is what will happen.

Facts or persuasion in health care decisions?

Posted on by

A thought-provoking perspective article in the NEJM 2012; 367: 1677-79 makes the argument that public health organizations/health professionals should be providing facts to inform people and not try to persuade or dissuade  them from screening for health conditions.

Examples of persuasion:

  • “The early warning signs of colon cancer: You feel great. You have a healthy appetite. You’re only 50.” (ad from New York Times Magazine from Memorial Sloan-Kettering Cancer Center)
  • “If you haven’t had a mammogram, you need more than your breasts examined.” (from American Cancer Society, 1970s)

The problem with persuasion messages is that they are ‘stripped of useful facts’ that indicate that screening can have harms as well as benefits.

For colon cancer, most 50-year-olds who feel great will not develop colon cancer.  The National Cancer Institute estimates that a 50-year-old has a 6 in 1000 chance of developing colon cancer in the next 10 years; the chance of death from colon cancer in that period 2 in 1000.  The advantages of colon screening would be to reduce that risk further: 67% with colonoscopy and 26% with sigmoidoscopy.  Screening, however, could result in bleeding, anesthetic complications, or perforation.

The tactics of persuasion rely on making people feel vulnerable and offering them hope.  Informed decisions allow appropriate expectations about the true benefits relative to harm.

Life in the balance (book)

Posted on by

While a recent blog (“There is More to Life Than Death”) referred to the complexity of  population-based medical decisions when relying solely on mortality, clearly mortality matters a lot.  In the same NEJM issue, an article highlights the difference that expanded Medicaid coverage has on mortality in adults (NEJM 2012; 367: 1025).

Traditionally, Medicaid has covered only low-income children, parents, pregnant women, and disabled persons.  In the past decade, several states expanded coverage to include nondisabled adults without dependent children.  This study examined these efforts in three states (New York, Arizona & Maine).  Then, effects on mortality were examined from 1997-2007 in these states as well as bordering states who did not expand coverage.  This time period allowed data to be analyzed five years prior to change and five years afterwards.

Findings:

  • By broadening eligibility requirements, there was a 25% increase in Medicaid coverage.
  • States with Medicaid expansions reduced all-cause mortality by 19.6 deaths per 100,000 adults; this was a relative reduction of 6.1%.
  • Mortality reductions were greatest for older adults (35-64), nonwhites, and residents of poorer counties.
  • The authors note that these changes do not prove causality.  However, there are implications for the affordable care act which allows expansion of eligibility up to 138% of the federal poverty level.

Despite apparent improvement in mortality, the cost, logistics and politics of expanding medical coverage remain unclear. In Georgia, the decisions on expanding medical coverage are quite controversial (see links below). My view: I think everyone in this country needs at least basic medical coverage.  This saves lives.

Deal rejects expansion of Medicaid | www.ajc.com

Expand Medicaid? – Blogs – Atlanta Journal-Constitution