Thursday, April 01, 2010

If they could only all be taken to Rampart Hospital

Trauma surgery for me is a whole lotta non-operative care interspersed with occasional surgery for things like a ruptured spleen. This is because I don't practice in a large urban center, and most of the trauma patients I care for have suffered blunt force injuries rather than penetrating ones. For most of the non-physician population, the idea of trauma surgery is heavily influenced by television --- shows like "E.R.," "Grey's Anatomy," and the like (I am asked at least weekly whether I enjoy these shows; I haven't seen a medical drama since "Emergency" ended its run in the 70's). Like other dramas, these shows try to maximize tension to keep the viewer interested; in trauma, there is nothing quite like a gunshot victim to achieve that goal. Lots of blood and screaming, along with an intense urgency to getting a patient off to the operating room.....where they are miraculously saved every time.

Every time. Except in real life, where some patients die of their gunshot wound (GSW). Sometimes, the reasons are obvious --- shot through the heart or the head with a fatal brain injury. Sometimes, the reasons are harder to understand ---- see a description of irreversible coagulopathy here. As for the rest, we don't have all of the answers, though not for a lack of searching.

Insurance Coverage Is Associated With Mortality After Gunshot Trauma is a recent retrospective study that is part of that searching. The trauma department at UCSF - East Bay Alameda County Medical Center in Oakland reviewed the records of 2,164 patients over a 10 year period who presented to their facility with a trauma activation for GSW. A few salient points from this report:

  • 92% of GSW victims were male
  • Average age of 28 (+/-9), with no difference between insured and uninsured patients
  • Injury severity scores (a system to "score" how badly a patient is injured) were similar between groups
  • ~25% were insured and less than 1% were on Medicare or Medicaid
  • Overall hospital mortality was 8%
  • There were no differences in mortality between Hispanic, Caucasian, and African-American patients
  • Uninsured patients had an odds ration for death of 2.2 in comparison to insured patients, despite access to the same level of care
This disparity is puzzling. From the authors' summary,
...despite similar injury severity, uninsured trauma patients were significantly more likely to die after admission for gunshot injury than insured patients. This difference could not be attributed to racial demographics or hospital resource use, and it held true even after adjusting for the effects of race, age, gender, and injury severity.
On possibility is that patients who died had more medical problems....except that insured patients were found to have more comorbid conditions than uninsured patients (17% versus 12%). In fact, because this was felt to represent an increased exposure to medical care on the part of insured patients, comorbid conditions were specifically excluded from the data analysis in this study.

In the end, the authors concluded that coverage most likely reflects social environment and the many social determinants of health. Social support networks, coping skills, and similar social factors likely affect outcomes after gunshot trauma..... The health burden of chronic social stress has been well studied and even physiologically quantified,and environmental stress is known to affect the health of young adults.

...Lack of insurance may be a reflection of the social environment and the many social determinants of health. Improving the social environment of patients affected by violent trauma is a potential intervention to improve mortality from gunshot trauma.
What do I think? Meh. Several things strike me at the same time.
  • It's hard to extrapolate the effect of varied social environments to medical outcomes. In fact, this is one of those areas where well-meaning researchers sometimes reach for non-quantifiable data to come up with answers that probably don't fully satisfy the questions they pose.
  • When such disparities in outcomes occur following full implementation of SocialistObamaCare, when "everyone will have coverage" --- and the disparities inevitably will occur --- how will we account for those differences?
  • Maybe the insured patients had caring family members praying for them, and more of the uninsured had support systems consisting of gang members who really didn't care if they made it out of the hospital walking or with a toe tag. This, I suppose, would play into the authors' feeling that social support networks are important; I just don't think we as a society do a good job of fostering this type of behavior.
  • Nothing good ever happens after midnight, like getting shot. In case you were wondering.
Just a little food for thought.