Monitoring Preload in the ICU: End-Diastolic Volume or Wedge?
ABSTRACT & COMMENTARY
Controversy persists over how best to monitor cardiac preload in the management of critically ill patients. For more than 20 years, the pulmonary artery occlusion (or wedge) pressure has been the clinical measurement of choice in gauging central volume status and the response to fluid therapy when noninvasive bedside assessment is considered to be insuffient or inaccurate. However, measurement of right ventricular ejection fraction and calculation of right ventricular end-diastolic volume index (RVEDVI) has been held by some intensivists to be superior to the wedge pressure in predicting the response to fluid challenge. Both wedge pressure and RVEDVI require a pulmonary artery catheter; to determine the latter, a special (more expensive) catheter with a rapid-response right ventricular thermister is necessary. This prospective study was designed to determine the correlation between RVEDVI and change in stroke volume in response to fluid challenge, to test the often-cited lower and upper limits of clinically acceptable RVEDVI, and to compare the wedge pressure to the RVEDVI as a predictor of hemodynamic response to fluid challenge in critically ill patients.
Wagner and Leatherman administered 36 volume challenges to 25 patients in two medical ICUs in whom rapid-response thermister-equipped pulmonary artery catheters had been inserted. Half of the patients had clinical sepsis, while the rest had a variety of acute medical illnesses such as ARDS, COPD exacerbation, and upper gastrointestinal bleeding. During volume challenges, saline or colloid was administered rapidly until the wedge pressure rose by at least 3 mmHg.
Twenty patients demonstrated an increase in stroke volume of at least 10% with fluid challenge (responders), while 16 did not (non-responders). Baseline wedge pressures were significantly different in these groups: 10.0 ± 3.4 mmHg in responders vs. 14.2 ± 3.6 mmHg in non-responders (P < 0.001), but there was no difference in baseline RVEDVI (105 ± 31 vs. 119 ± 33 mL/m2 [P = 0.22]). The relationship between wedge pressure and change in stroke volume (r = 0.58) was stronger than that between RVEDVI and stroke volume change (r = 0.44). Further, a positive response to fluid challenge occurred in four of the nine cases in which initial RVEDVI exceeded 138 mL/m2, a threshold value above which a positive response has been considered to be unlikely. Wagner and Leatherman conclude that wedge pressure is superior to RVEDVI as an index of preload and predictor of the hemodynamic response to rapid fluid challenge. (Wagner JG, Leatherman JW. Chest 1998;113(4):1048-1054.)
COMMENT BY DAVID J. PIERSON, MD
Because hypotension, oliguria, and tachycardia are commonly due to relative hypovolemia in critically ill patients, it is commonplace to administer fluid challenges empirically in assessment and management. However, in patients who are euvolemic or fluid overloaded, the administration of additional volume can lead to pulmonary edema, arrhythmias, and other potentially life-threatening consequences. A reliable bedside predictor of fluid responsiveness would thus be desirable.
Two retrospective studies have concluded that RVEDVI, as estimated using a rapid-response thermodilution pulmonary artery catheter, is a useful predictor of whether a given patient will benefit hemodynamically from fluid administration (Reuse C, et al. Chest 1990;98:1450-1454; Diebel LN, et al. Arch Surg 1992;127:817-822). Both studies found that patients with the highest RVEDVI values failed to respond to fluids; these cut-off values were 140 mL/m2 (Reuse et al) and 138 mL/m2 (Diebel et al), respectively. The latter study also found that all patients with RVEDVI less than 90 mL/m2 responded positively to fluid challenge. In both of these studies, the wedge pressure was less predictive of a positive response to fluids than was RVEDVI.
Why are the results of the present study different from those of these and other previous studies? As Wagner and Leatherman discuss, there may have been some technical differences in the measurements. They injected the indicator solution without regard for the respiratory cycle, whereas the previous investigators synchronized the injection with end-inspiration, potentially introducing a systematic error. Perhaps, more importantly, in my opinion, the studies may have been done using sufficiently different patient populations and management approaches that direct comparison of their findings becomes problematic. The patients studied by Wagner and Leatherman were medical patients, not surgical patients or victims of multiple trauma. They may have been relatively more hypovolemic than those in the earlier studies (their wedge pressures tended to be lower), and, thus, on somewhat different portions of their Starling pressure-volume curves. Only two patients in the present study had ARDS, so that there may also have been substantial differences in the amount of PEEP used.
In any case, neither measure of central volume status was all that precise, and despite the statistically significant correlations, values for both wedge pressure and RVEDVI showed substantial splay when plotted against changes in stroke volume. Thus, the controversy over how best to predict a favorable response to fluid challenge and what individual predictor to use is likely to persist. The most pragmatic approach for determining the volume needs of many individual patients will likely remain the clinical assessment of cardiac performance before and after the empiric administration of a bolus of fluid, regardless of which technique or hemodynamic index is used.
You have reached your article limit for the month. Subscribe now to access this article plus other member-only content.
- Award-winning Medical Content
- Latest Advances & Development in Medicine
- Unbiased Content