|
|
The Incidence of Major Morbidity in Critically Ill Patients Managed with Pulmonary Artery Catheters: A Meta-analysis Ivanov R, Allen J, Calvin JE. Crit Care Med 2000; 28(3): 615-619
Reviewers: John G. Augoustides, MD; Albert
T. Cheung, MD
Background: There is still doubt whether using the pulmonary artery catheter (PAC) improves patient outcome despite more than 25 years of clinical experience with the device from around the world. This report attempted to address this important question by performing a meta-analysis of PAC clinical trials with morbidity as the outcome end-point. The authors have performed previously a similar meta-analysis of PAC clinical trials using mortality as the outcome end-point, but that meta-analysis failed to show a significant trend towards lower mortality in patients managed with a PAC (New Horiz 1999;5:268-276). Methods: A literature search beginning with studies published in 1970 yielded 12 randomized, controlled clinical PAC trials with data on morbidity. Two independent reviewers scored each study in terms of design, data collection and analysis using the standard method described by Chalmers (Control Clin Trials 1981;2:31-49). An average score was then calculated for each individual study. New organ system dysfunction was chosen as morbidity end-points. These morbidity end-points were defined using standard criteria based on clinical and laboratory measurements for the following organ systems: pulmonary, cardiovascular, renal, hepatic, hematologic , gastrointestinal and neurologic. The development of sepsis based on standard criteria was also included as a morbidity end-point. Complications that did not produce organ dysfunction or sepsis according to the defined criteria were not included in the analysis, but were described in detail. Differences in morbidity in patients managed using a PAC (treatment group) versus patients managed without a PAC (control group) were then analyzed using 2X2 contingency tables. A random effects model was used to account for variability between the individual studies. Results: The meta-analysis included 1,610 patients pooled from 12 studies. The incidence of morbidity was 63% in the PAC treatment group and 74% in the control group. The relative risk of morbidity was significantly less in the PAC treatment group (relative risk ratio = 0.78, 95% confidence interval: 0.65-0.94, p=0.017). The PAC treatment group had a mean protective effect of 21.9% for risk of morbidity. Patient acuity, study quality score, when the study was performed, study comparison type and intensive care unit type were not significant predictors for the risk ratio of morbidity. The relative risk ratio for morbidity did not change significantly across the studies suggesting that the result was not dependent upon a single influential study. In addition, there was no report of patient morbidity caused directly by PAC insertion in the 1,610 patients included in the analysis. Conclusion: The authors concluded that major morbidity defined as new organ system dysfunction or sepsis was reduced significantly in critically ill patients managed using a PAC. Comments: The controversy surrounding the clinical utility of the PAC study was re-ignited in 1996 when Connors, et al reported the results of a large clinical trial in JAMA that found a higher mortality in intensive care unit patients who had a PAC (JAMA 1996; 276:889-897). The Connors study sparked a major clinical controversy about the future of the PAC in clinical practice not only because PAC use did not improve outcome, but also because PAC use was believed to have increased the costs of care without proven benefit. However, the conclusion generated by the Connor study has been questioned because the study was retrospective and the comparison between treatment and control groups was not randomized and used a controversial scheme for identifying historical case-matched controls. Although the best approach for addressing the question of whether there is actual benefit associated with the use of the PAC is to conduct a large multi-center prospective, randomized controlled clinical trial, the logistic and practical difficulties inherent in such an ambitious project have yet to be solved. Until a prospective trial has been accomplished, it is reasonable to investigate the clinical utility of the PAC by analyzing the results of previous trials using meta-analysis. Because a meta-analysis is only as good as the original studies and published data, the authors selected only randomized controlled clinical trials with a sound study design and chose to analyze only clinically important end-points that were defined by strict clinical and laboratory criteria. The inherent limitations of the meta-analysis include the quality of data reporting across the studies used for analysis. The authors acknowledge this potential problem as an entity that defies objective assessment. The absence of complications in 1,610 PAC insertions was concerning given the known incidence of PAC complications including catheter-related sepsis. The absence of PAC complications cannot be entirely explained by the involvement of experienced clinical investigators who inserted the catheters in the studies. Another design limitation was the inclusion of only published randomized controlled PAC trials in the meta-analysis. Selecting only published trials that tend to report favorable findings may have introduced a sampling bias. This limitation was also acknowledged, but it was argued that it was preferable to sacrifice sampling size in favor of using studies that were well designed and generated statistically sound data. The results of this meta-analysis falls short of providing a definitive answer to the controversy surrounding the use of the PAC. Nevertheless, the study does provide additional evidence to back the continued use of the PAC that can be cited by clinicians who still believe that the data provided by the PAC is important and has a favorable impact on managing the critically ill patient.
Links: Vasopressin in the treatment of refractory hypotension Evidence-based
Pharmacologic Therapy During Cardiac Arrest
|