Date of Award

11-3-2006

Document Type

Thesis

Degree Name

Medical Doctor (MD)

First Advisor

Lewis Kaplan

Abstract

Current understanding of acid-base disturbances in trauma patients may rely on incomplete data and flawed methods of analysis. We examine existing manuscripts to determine if they contain sufficient detail to completely understand the nature of acidosis following injury. We also study the etiologies and history of trauma-related acidosis, and compare two resuscitation schemes and acid-base interpretation methods the physicochemical approach (PC) or the Henderson-Hasselbalch (HH) approach. Phase 1 included a Medline search for trauma and acid-base manuscripts (1994 2005). Manuscripts were categorized by type of data, data completeness, and linkage of acid-base abnormalities to outcome. Phase 2 was a prospective observational analysis of two groups of trauma patients requiring SICU care. Group 1 was cared for by an Intensivist utilizing the PC approach and Group 2 was cared for by other traumatologists utilizing the HH approach. Demographic and outcome data were derived from the trauma registry. Laboratory data were used to calculate the anion gap, standard base excess, strong ion differences, and strong ion gap. Physician analysis of each patients acid-base disturbance(s) upon admission to the ICU were recorded, as were the resultant interventions. Days until normalization of electrolyte and/or blood gas values were tracked. The Intensivist reviewed Group 2 acid-base analyses and interventions. An external reviewer (blinded) examined patient data to render independent acid-base analyses and critique intervention appropriateness. The reviewer listed potential morbidities resulting from the intervention(s). Unblinded, the reviewer analyzed the Intensivists critique of Group 2 acid-base determination and intervention. Discrepancies between the Intensivist, the Group 2 physicians and the reviewer were collated. The Phase I search yielded 1286 manuscripts, 238 of which were evaluable. 56% were animal, 22% retrospective human, 21% case series, and 1% prospective randomized controlled trials. 59% indicated the resuscitation fluid used, 44% the volumes used, 55% provided laboratory data before and after resuscitation, but only 25% linked acid-base abnormalities to outcome. 120 manuscripts presented 1 of 4 essential data points (chloride, lactate, fixed acids, unmeasured ions), 12 presented 2 of 4, 2 presented 3 of 4, but only 5 presented all four data points. During Phase 2, there were no differences in mortality or ISS between groups; there were more operations in Group 2. The Intensivist disagreed with 92% of Group 2 evaluations of acid-base disturbance(s), 98% of Group 2 fluid prescription, and 11% of Group 2 ventilator interventions. The reviewer agreed with these analyses with the exception of one ventilator prescription. Group 1 days until normalization (3.31 ± 3.42) were significantly shorter than Group 2 (8.29 ± 7.35; p < 0.005). Group 1 hospital length of stay (days) was significantly shorter (8.18 ± 11.50 vs. 15.57 ± 19.26, p = 0.02). We conclude that existing literature rarely provides sufficient data for a physicochemical analysis of acid-base balance. A standard reporting format for acid-base analyses would create a useful parity between studies. The physicochemical approach allows for a more precise determination of acid-base disturbances, and guides resuscitation and fluid prescriptions using physiologic principles. This approach may lead to improved resource utilization.

Comments

This thesis is restricted to Yale network users only. This thesis is permanently embargoed from public release.

Download

Share

COinS