Date of Award

January 2016

Document Type

Open Access Thesis

Degree Name

Medical Doctor (MD)

Department

Medicine

First Advisor

Michael L. Diluna

Abstract

Traumatic fractures of the upper cervical spine are a diverse group of injuries with high morbidity and mortality. Guidelines for treatment of these injuries have been based off limited medical literature; the factors associated with the management decisions and the outcomes associated with those decisions have not been well described. In particular, recent literature supports the idea that nonclinical factors such as race and insurance may affect treatment decisions for all spine fractures.

The electronic medical record was queried for trauma patients admitted to Yale New Haven Hospital diagnosed with an upper cervical spine fracture using relevant ICD-9 codes. Information on patient demographics, injury details, treatment, procedures, and outcome were gathered. Descriptive statistics and univariate analysis were used to describe the population and the risk factors associated with each treatment cohort and various outcomes.

Of 239 patient records reviewed by the time of publication, 218 had definitive evidence of upper cervical spine fracture. Overall in-hospital mortality rate was 11%. Mean age of injury was 69, with median age of injury 77 years. 85.8% (n= 194) patients were white. The primary payer insurance was governmental for 150 patients; 41 patients had private or commercial coverage (18.8%), while 20 patients had no insurance coverage at admission (9.2%). White patients were more likely to undergo initial treatment in rigid collar (OR = 0.30, p= 0.002) and were also more likely to receive collar as definitive treatment (OR= 0.30, p= 0.002). Nonwhite patients were 3 times more likely to receive initial treatment in a halo orthosis (p= 0.005) and were also 3 times more likely to receive halo orthosis as definitive treatment (p= 0.012). Race was not associated with initial surgical treatment (OR= 1.88, p= 0.246) or with definitive surgical treatment (OR= 2.01; p= 0.110). Patients without insurance had a lower likelihood of receiving CT upon admission (OR= 0.31, p= 0.05) while patients with government insurance had a significantly lower chance of receiving a CTA, MR or MRA during admission compared to those with private insurance (ORCTA= 0.30, p= 0.001; ORMR= 0.40, p= 0.023; ORMRA= 0.37; p= 0.01). The only significant association between insurance and treatment was the increased likelihood of patients without insurance to receive definitive treatment in halo orthosis (OR= 3.79, p= 0.010). Insurance was not associated with risk of death during admission. None of the patient factors or injury factors assessed had an effect on the likelihood of receiving surgery as primary treatment; the only patient factor associated with an increased risk of definitive surgical treatment was CCI. Race and insurance were not associated with death during admission. Management in collar or by surgery was associated with increased death during admission, as was C2 Type and dens Type 2 fracture. Race did not affect risk of nonunion but decreased likelihood of delayed healing during follow-up.

Being male, nonwhite, under age 65 and being either uninsured or having government insurance significantly increased the likelihood of receiving halo orthosis as definitive management for UCS fracture. Being uninsured or having government insurance decreases the likelihood of receiving specialized imaging during admission for upper cervical spine fracture. Further analysis using multivariate techniques may yield more nuanced investigation of nonclinical factors in treatment and outcomes of upper cervical spine fracture at Yale.

Comments

This is an Open Access Thesis.

Open Access

This Article is Open Access

Share

COinS