Safe Prehospital Treatment of Chest Pain Without On-line Medical Control

Abstract

Purpose:To test the accuracy of pulse oximetry and capnometry aboard a moving ambulance in a cardiogenic shock model. Methods: Five dogs were anesthetized with pentobarbital, intubated, and instrumented. Stepwise worsening cardiogenic shock was induced with esmolol until cardiac output dropped 75%, then reversed in steps. An arterial blood gas specimen and hemodynamic measurements were taken at each step. Weeks later, the experiment was repeated aboard a moving ambulance on four dogs; shock was induced quickly and reversed slowly. O2 saturation was estimated by pulse oximetry and pCO 2 by end-tidal CO 2 Differences between the noninvasive estimate and the blood gas measurements (measurement error) were compared statistically. Results: Pulse oximetry and capnometry usually were accurate, but some highly erroneous readings occurred. Pulse oximetry readings sometimes were unavailable, especially during profound shock. Mean error SD 50th % 75% 100% % Obtained O2Sat 7 9 2 8 47 87 pCO2 10Hypothesis: Patients with uncomplicated chest pain can be treated safely and transported by paramedics without on-line [direct] medical control (OLMC). Methods: A retrospective chart review of all non-traumatic patients with chest pain treated in a combined rural and suburban emergency medical system over a two-year period (December 1990-November 1992) was conducted. Prior to 1 November 1991, OLMC was mandatory for all patients with chest pain. Beginning 1 November 1991, if a patient had pain resolution spontaneously, with the administration of oxygen, or following a single dose of nitroglycerin, OLMC was at the discretion of the paramedic. Using the above criteria for inclusion, two study groups were defined, before and after protocol change (groups I and II respectively). The groups were compared using the following study parameters: scene time, transport time, time to administration of first nitroglycerin, number of vital signs completed, oxygen use, IV access, and ECG monitoring. Continuous and categorical variables were analyzed by Rest and x 2 respectively. Results: Of 303 non-traumatic, adult chest pain patients, 55 met inclusion criteria for group I and 39 for group II. The majority of patients were excluded due to persistent pain requiring OLMC. The sample size yielded a power of 80% to detect a difference of 20% at a = 0.05. No statistical difference was found in any of the study parameters. Conclusions: Uncomplicated, adult, chest pain patients can be treated appropriately by paramedics without OLMC.Errors in O 2 saturation estimation increased by about 5 torr when made aboard the ambulance (p = .04); increased errors in pCO 2 estimation (<1 torr) were insignificant clinically. Pulse oximetry also was compromised by decreasing cardiac output (p= .01) and a hysteresis effect (p= .05). The pCO 2 estimation became less accurate with decreasing blood pressure (p = .02). Partial data were collected on two dogs: one had an intercurrent lymphoma, the other died following an infusion pump malfunction. Conclusion: Pulse oximetry and ...