A comparison of post-injury infectious outcomes in elderly patients with Medicare vs private insurance

Theodore Delmonico; Andrew Stephen; Charles Adams, Jr.; Tareq Kheirbek; David Tobin Harrington; Stephanie Lueckel; William Cioffi; Daithi Heffernan

Health care insurance status is associated with differences in trauma related outcomes. We previously demonstrated that under- or un-insured young trauma patients were less likely to be diagnosed with comorbidities and more likely to develop infectious complications. Diagnosing and managing multiple comorbidities in a geriatric patient are both time intensive and costly, resources often lacking among patients with government based insurance. Geriatric patients with no or few comorbidities may be truly healthy or may lack maximal diagnoses and management. 

Given constraints upon Medicare/Medicaid (MC) providers and patients we hypothesize that geriatric trauma patients with MC insurance will have increased rates of infections.

A retrospective chart review all admitted geriatric blunt trauma patients(>/=65 years old) over a 10 year period. Patients were divided into MC or Private/Commercial (PI) insurance. All injuries, pre-trauma co-morbidities, hospital course, infections, and mortality were recorded. Infections were diagnosed via either culture based (UTI, catheter, pneumonia confirmed via alveolar lavage) or CDC clinical criteria (eg surgical site infections).

Of 7,319 geriatric trauma patients, 85.9% were MC. MC patients were older(81.1 vs 77.5 years;p<0.01), less likely male(38.8% vs 46.9%; p<0.01) and had lower ISS (10.1 vs 11.2;p<0.01). There was no difference in Head AIS. Patients with MC had higher rates of infections(11.9% vs 9.2%;p=0.012). Hypertension was the most common diagnosis among both groups. Among patients with no more than one comorbidity, MC patients had higher rates of infection(10.6% vs 7.7%;p=0.007). Adjusting for age, gender and ISS, MC patients had greater risk of infection (OR-1.32(95%CI=1.1-1.7)). Among patients with at least 2 comorbidities there was no difference in rates or risk of infection. Patients with dementia had the highest rates of infection with no difference between MC and PI(21.2% vs 21.3%;p=0.97). Among PI patients, increasing numbers of infections did not affect risk of death, whereas among MC patients, each added infection among an individual patient significantly increased the risk of death(p<0.006).

Geriatric MC patients have higher rates of infections. We speculate that inherent difficulties exist in providing maximal diagnostic care among patients with MC health coverage. Future MC policies and decision making need to focus on optimizing diagnoses and health maintenance services for the elderly MC population.