Outcomes After Pediatric Trauma in Alaska Depend On the Distance From Surgical Care and Surgical Injury Pattern

Sunday, October 27, 2013: 11:12 AM
Windermere Ballroom W (Hyatt Regency Orlando, formerly the Peabody)
Christopher W. Snyder, MD, MSPH1, J. Brent Roaten, MD, PhD1, Frank Sacco, MD2, Oliver J. Muensterer, MD, PhD3 and Shawn David Safford, MD4, (1)Surgery, The Children's Hospital at Providence, Anchorage, AK, (2)Alaska Native Medical Center, Anchorage, AK, (3)Pediatric Surgery, Weill Cornell Medical Center, New York, NY, (4)Surgery, National Naval Medical Center, North Potomac, MD


Pediatric trauma care in Alaska is limited by the great distances that children must travel in order to receive surgical treatment.  This provides a unique opportunity to study the effects of distance from surgical care on outcomes after injury in children. 


We performed a retrospective cohort study using State of Alaska Trauma Registry data for all children age 0 to 17 years who sustained non-trivial (Injury Severity Score >=1) traumatic injury between January 2001 and December 2011. Surgical diagnoses (diagnoses expected to require at least evaluation by a surgeon) were defined based on ICD-9 codes, then further classified into three categories to more precisely define the surgical injury pattern: isolated CNS, isolated non-CNS, and combined CNS and non-CNS. The straight-line distance from the zip code where the injury occurred to the nearest zip code with a practicing general surgeon was calculated. The primary outcome of interest was a composite of either death or permanent disability. Logistic regression was used to evaluate the effects of surgical injury pattern and distance from surgical care on outcome, with adjustment for relevant covariates and interactions between variables.


A total of 5,547 records were included. Significant regional differences were found with respect to ethnicity, community type, injury severity, distance to surgical care, and outcome. On unadjusted bivariate analysis, the surgical injury pattern was strongly associated with death or permanent disability. Non-CNS, isolated CNS, and combined patterns carried risks of 4.4%, 10.8%, and 43.7% for unfavorable outcome, respectively, all of which were significantly higher (p<0.05) than the risk with no surgical diagnoses (1.8%). On logistic regression modeling with adjustment for relevant covariates, there was a strong interaction between surgical injury and distance to nearest general surgeon (p-value for interaction = 0.003). Isolated injury patterns occurring within the same zip code as a general surgeon (i.e. modeled distance of zero miles) were associated with lower risk of unfavorable outcome (Odds Ratio 0.42, 95% Confidence Interval 0.23-0.76), but the risk of unfavorable outcome increased exponentially with increasing distances from surgical care. At 400 miles the risk of unfavorable outcome was significantly greater (OR 8.8, 95% CI 1.7-44.4). Patients with combined CNS and non-CNS injuries followed the same pattern, but had higher risks of unfavorable outcome at all distances. 


Long distance from surgical care is a significant predictor of unfavorable outcome after injury in children. Although surgical injury patterns tend to carry a higher risk of unfavorable outcome, their effects are mitigated by close proximity to surgical care. These findings have important implications for delivery of pediatric trauma care in Alaska and other geographic areas that are impacted by great distances to trauma care.