Monday, January 05, 2009

VoTE for Prophylaxis

"Doc, don't hurt me!" I hear this not infrequently, as I am sure most surgeons do. Patients don't like pain, but some equate the commonly stated aphorism "Primum non nocere" with "don't hurt me." This would appear to be a bit of an oxymoron when dealing with surgery. It's kind of hard to not hurt someone when performing an operation --- let's face it, everything I do hurts --- so we surgeons want to make sure that the proper translation is used:

First, not to harm (or, more commonly, First, do no harm)

Sometimes, we do indeed harm patients. This is done mostly in the setting of trying to help them, but certainly complications and bad outcomes occur. Taken as a whole, "medicine" has gradually advanced, and prior harmful practices have been replaced with better ones with less potential for harm. Sometimes, the simple process of having an operation is potentially harmful, with significant changes in the immune system, cardiovascular system, hormonal stress response, and inflammatory mediator release having some degree of harmful effect ---- these do not always result in serious issues, but they can. Another potentially harmful side-effect of surgery is the risk of development of deep vein thrombosis and pulmonary embolism, which I have written about before. While not always preventable, adequate prophylaxis with certain types of major surgery significantly reduces the risk of developing postoperative venous thromboembolism (VTE) ...... but "adequate prophylaxis" is somewhat of a moving target.

(Please remember, when I am using the word prophylaxis, I am not alluding to what Frank Oz gave back to John Belushi in "The Blues Brothers.")

A pair of recent articles focusing on the risks of VTE with major abdominal surgery are interesting for what they tell us about prophylaxis. The first, entitled Postoperative Venous Thromboembolism Rates Vary Significantly After Different Types of Major Abdominal Surgery, is a retrospective review of >375,000 patients using the Nationwide Inpatient Sample.

Methods Retrospective analysis of the Nationwide Inpatient Sample (2001–2005) was conducted. Eight surgeries were identified: bariatric surgery, colorectal surgery, esophagectomy, gastrectomy, hepatectomy, nephrectomy, pancreatectomy, splenectomy. Age <>
Results Patients, 375,748, were identified, 5,773 (1.54%) with VTE. Overall death rate was 3.97%, but 13.34% after VTE. Unadjusted rate (0.35%) and adjusted risk for VTE were lowest among bariatric patients. On multivariate analysis, highest risk for VTE was splenectomy (odds ratio 2.69, 95% CI 2.03–3.56). Odds ratio of in-hospital mortality following VTE was 1.84 (1.65–2.05), associated with excess stay of 10.88days and $9,612 excess charges, translating into $55 million/year nationwide.

Conclusion Highest risk for VTE was associated with splenectomy, lowest risk with bariatric surgery. Since bariatric patients are known to have greater risk for this complication, these findings may reflect better awareness/prophylaxis. Further studies are necessary to quantify effect of best-practice guidelines on prevention of this costly complication.

That's a very interesting set of data to pore over, with a few surprises. First of all, the overall risk of developing VTE with these selected major abdominal operations was low at 1.54%. The most common genetic cause of venous thrombosis, Factor V Leiden (Factor V mutation) is present in 3% of the general population (estimated 2-7%, with only about 0.1% of the population being homozygotes). This study did not specifically address whether the patients with postoperative VTE were screened for this or other known genetic causes of venous thrombosis, and did not address other potential risk factors (oral contraceptive use, estrogen therapy, obesity, diabetes, venous stasis from immobility).

Just as interesting to me is the fact that patients undergoing bariatric surgery (typically gastric bypass surgery) had the lowest risk of development of VTE. This is counterintuitive, as these patients have a higher risk of VTE due to their obesity, frequent associated diabetes and venous stasis disease, and relative immobility. What gives? Did the numbers stop making sense after a few months of staring at reams and reams of data? Well, no. It appears that bariatric surgeons, knowing that their patients are at a higher risk for VTE, were more apt to provide patients VTE prophylaxis (subcutaneous Heparin or Lovenox). In the body of the article, we find this tidbit of information:
"While the ENDORSE trial recently found that only 58.5% of all surgical patients at risk for VTE received American College of Chest Physicians (ACCP)-recommended VTE prophylaxis. Wu and Barba showed that more than 95% of bariatric surgeons regularly adhere to thromboprophylaxis guidelines."
So, chicken or egg? Are bariatric surgical patients truly at a lower risk for VTE, or are their surgeons simply better at providing appropriate prophylaxis?

The next article looks only at VTE in bariatric surgery --- The effect of extended post-discharge chemical thromboprophylaxis on venous thromboembolism rates after bariatric surgery: a prospective comparison trial. Basically, these authors are asking the question "would it be better to extend the length of VTE prophylaxis beyond a bariatric surgical patient's hospitalization?"

Methods 308 consecutive patients who underwent BS between 2003 and 2007 and who had >1 month of follow-up were included. In-hospital-only VTE prophylaxis (group A), or extended 10-day ETP (group B) was used in 132 and 176 patients, respectively. All patients underwent bilateral lower extremity venous Doppler studies (BLEVDS) prior to discharge. Primary endpoint was the incidence of VTE within 30 days postoperatively. VTE was defined as a clinically evident deep vein thrombosis or pulmonary embolism documented by positive BLEVDS, or computed chest tomography. The primary safety endpoint was bleeding associated with ≥2 g/dL decrease in hemoglobin compared with baseline, transfusion or reoperation.

Results The incidence of VTE was 1.9% (6/308); 66.6% (4/6) of cases occurred after cessation of thromboprophylaxis. There were no deaths in either group. With the exception of percentage open surgical approach (A: 3% versus B: 0%, p = 0.03), percentage conversions (A: 0 versus B: 3.8%, p = 0.01), and hospital stay (A: 3 versus B: 2.2 days, p < style="font-weight: bold;">VTE rate was significantly higher in group A (A: 4.5% versus B: 0%, p = 0.006). Although morbidity was higher in group A (A: 12.1% versus B: 1.1%, p = 0.02).

Conclusions ETP is safe and effective in reducing the incidence of VTE as compared with in-hospital thromboprophylaxis only.
Once again, we see that the overall incidence of VTE was small -- 1.9% in this study. However, 2/3 occurred after cessation of propylaxis (these four patients developed VTE on postop days 12, 20, 26, and 30). Unfortunately, because the study size was small, there were some significant differences in VTE risk in the two groups, which could have something to do with the difference in outcomes. A larger study would be very helpful, but it is a good thing that more folks are thinking about extending the time frame for VTE prophylaxis.

I think these types of studies are extremely helpful for clinicians. I think they would be very helpful to policy-makers as well, who are pushing the idea that VTE should be a "never event," but I'm not so sure they are as interested in gathering relevant data as in holding on to their cash.