Ultrasound and Computed Tomography Imaging Technologies for Nutrition Assessment in Surgical and Critical Care Patient Populations

Computed Tomography-Measured Psoas Muscle Density as a Predictive Factor for Hypophosphatemia Associated With Refeeding

BACKGROUND

Initiation of parenteral nutrition (PN) after a period of starvation can be complicated by refeeding syndrome (RFS). RFS is associated with electrolyte abnormalities including hypomagnesemia, hyponatremia, and hypophosphatemia. Risk factors include recent weight loss, low body mass index, and electrolyte deficiencies; however, these associations are not strong. We hypothesized that a validated measure of nutrition risk, computed tomography (CT)-measured psoas muscle density, can be used to predict the development of hypophosphatemia associated with RFS.

METHODS

A retrospective analysis of surgical patients initiated on PN with an abdominal CT scan within the past 3 months was conducted. CT-measured psoas muscle density was assessed as a predictive variable for the development of electrolyte abnormalities. Daily electrolyte and clinical outcome measures were recorded.

RESULTS

One hundred nine patients were stratified based on Hounsfield unit average calculation (HUAC). The lowest 25th percentile of patients had HUAC <25. Low HUAC was associated with a significant percent decrease in phosphate levels from baseline to PN day 3 (P < .01) and significant difference in serum phosphate value on PN day 3 (P < .01). The low muscle density quartile also experienced longer days on the mechanical ventilator (P = .01) compared with patients with a higher psoas muscle density.

CONCLUSION

Psoas muscle density predicted the development of hypophosphatemia in patients initiated on PN. This measurement may aid in identifying patients at highest risk of experiencing RFS. A mean psoas HU <25 may prompt additional precautions, including additional phosphate replacement and slower initiation of PN.

Computed tomography measured psoas density predicts outcomes in trauma

BACKGROUND

Age-related loss of muscle mass and function (sarcopenia) is linked to poor outcomes after operation and trauma. Here we evaluate computed tomography measured psoas muscle density and area using quick and simple tools available to the bedside clinician. We hypothesize these measures will predict poor outcomes after blunt traumatic injury.

METHODS

We conducted a retrospective cohort study of patients ages ≥45 years in the Ohio State University Trauma Registry in 2008 that received a computed tomographic scan of the abdomen and/or pelvis with intravenous contrast. Psoas Index and Hounsfield unit average calculation were measured at the L3 level. In the study, 90-day mortality, complication, duration of stay ≥7 days, and dependent discharge were compared with Psoas Index and Hounsfield unit average calculation.

RESULTS

In the study, 151 patients met the inclusion criteria. Patients were stratified into interquartile ranges based either on Psoas Index or Hounsfield unit average calculation values. After adjustment with sex-specific cutoffs, the lowest interquartile range of Psoas Index was associated with 90-day mortality (relative risk [RR] 5.95, P < .008), but did not reach significance in other outcomes. The lowest interquartile range of Hounsfield unit average calculation was associated with 90-day mortality (RR 5.95, P < .008), duration of stay ≥ 7 days (RR 1.63, P = .048), complication risk (RR 2.30, P = .002), and dependent discharge 2.14, P = .015).

CONCLUSION

Psoas muscle density is a significant predictor of poor outcomes after traumatic injury. This objective, quick, and readily available measure of sarcopenia can identify patients requiring aggressive nutritional and physical therapy to improve prognosis, prevent recurrent traumatic injury, and aid in discharge planning.