The Effect of Prompt Physician Visits on ICU Mortality and Cost
The Effect of Prompt Physician Visits on ICU Mortality and Cost
Abstract & Commentary
James E. McFeely, MD, Medical Director Critical Care Units, Alta Bates Summit Medical Center, Berkeley, CA, is Associate Editor for Critical Care Alert
Synopsis: A retrospective study comparing patients seen within the first 6 hours of ICU admission to those seen later showed that patients seen earlier had a decreased risk of death, shorter hospital stays, and similar direct variable costs.
Source: Engoren M. Crit Care Med. 2005;33:727-732.
This observational study from a university-affiliated, urban, tertiary hospital in Toledo, OH, retrospectively reviewed the records of 840 patients admitted to various ICUs to determine if the time to first visit by a physician had an effect on a number of clinically relevant outcomes. The hospital operates with an open ICU structure, with some patients covered by housestaff and others cared for directly by attending physicians. Data collected included APACHE0 (scores calculated from data available at the time of admission), APACHE II scores, many physiologic variables, location prior to ICU admission, ICU and hospital length of stay, and hospital outcome. Direct variable costs were also calculated. Patients were divided into 2 groups: the "Prompt" group, defined as patients seen by a physician within 6 hours of ICU admission, and the "Delayed" group, seen more than 6 hours after admission.
To help separate out the effects of severity of illness and other factors that might affect a physician’s decision to see the patient promptly, a propensity score was calculated to predict the likelihood of belonging to either the Prompt or Delayed group. This score was then used in multivariate modeling of mortality and used to find patients with similar propensity scores in the 2 treatment groups for purposes of comparison.
Median time to first physician visit was 6 hours. Patients in the prompt group differed from those seen later in several ways. They were seen in different ICUs. Patients in the prompt group were more likely to be seen by housestaff and more likely to be admitted during the day; they had higher APACHE II and APACHE0 scores, and higher direct variable costs. Despite these differences, these patients had a lower mortality ratio (55% vs 88%) as predicted by APACHE II. In the subgroup matched by propensity scores, the prompt group had shorter hospital stays (11 d vs 13 d), similar direct variable costs and similar mortality rates. By binary logistic regression, higher APACHE0, score, older age, mechanical ventilation on arrival, and longer time to being seen by a physician were predictors of hospital mortality. Each 1-hour physician delay in initially seeing the patient was associated with a 1.6% increased risk of hospital death.
Comment by James E. Mcfeely, MD
This article contributes to a body of work published in the last few years studying the effect of staffing models and speed of intervention on ICU outcomes. Most of that work has focused on the organization of the ICU (closed vs open models) or on treatments for specific diseases such as early goal-directed therapy for sepsis, thrombolytics for acute myocardial infarction, and treatment for stroke. The overriding theme of that work is that faster is better. This paper answers the question more generally, by confirming that patients seen by a physician within 6 hours of ICU admission do better than those seen less promptly. The study was performed in a hospital with an open ICU structure, and included a wide variety of admitting diagnoses. Even after controlling for initial severity of illness and other factors using sophisticated statistical methods, researchers still found that time to initial physician visit was a significant predictor of hospital mortality.
As with all good research, this paper raises as many questions as it answers. Can outcomes be improved by seeing the patient within the first hour as with other specific diseases? Does it matter if the initial visit is by an attending physician or by a member of the housestaff? Would outcomes be different comparing an intensivist to a generalist making the first visit? Although this study found no difference based on the type of physician that made the initial visit (good news for those of you with housestaff), the question is difficult to answer in a randomized fashion, and further observational studies from a greater variety of institutions are required to gain better insight into these and similar issues.
Despite the need for further research, however, this paper should cause ICU medical directors to reconsider the practice of allowing ER staff to tuck in patients, and probably should prompt a change in policy to require onsite physician evaluation within a few hours of admission to the ICU.
This observational study from a university-affiliated, urban, tertiary hospital in Toledo, OH, retrospectively reviewed the records of 840 patients admitted to various ICUs to determine if the time to first visit by a physician had an effect on a number of clinically relevant outcomes.Subscribe Now for Access
You have reached your article limit for the month. We hope you found our articles both enjoyable and insightful. For information on new subscriptions, product trials, alternative billing arrangements or group and site discounts please call 800-688-2421. We look forward to having you as a long-term member of the Relias Media community.