Larry Lewis, M.D.
Project Overview:
Our objective is to improve the health outcomes and resource use of elderly patients presenting to the Emergency Department (ED) with complaints of abdominal pain. Abdominal pain is the most common chief complaint of patients presenting to EDs in the U.S., accounting for almost 5,000,000 visits annually (ACEP, 1994). It is also a common chief complaint among seniors, associated with almost 10% of ED visits among patients 60 years of age or older (Banet, Personal Communication). Abdominal pain presents unique challenges in the elderly. Their presentation of abdominal pathology often differs significantly from the classic textbook descriptions, as well they have a larger differential diagnosis of potentially life threatening diseases than younger counterparts, which must be considered and ruled out. (Balsano et al., 1990; Kauvar et al., 1993; Lee et al., 2000). As a result, significant resources are used in the diagnostic evaluation of these patients. Despite the liberal use of evaluative testing, there continues to be a high rate of mis-triage, mis-diagnosis and associated adverse outcomes including death (Hawthorn, 1992).
We plan to refine the current approach to the diagnostic work-up for seniors with abdominal pain as well as to develop a risk stratification schema that will enable clinicians, based upon the clinical presentation, to reliably predict which patients require a more extensive evaluation and/or hospitalization. The Abdominal Pain Evaluation Tool will allow us to determine the weight given by physicians to the various combinations of historical and physical examination variables in their diagnostic and therapeutic decision making process. We will then correlate various combinations of presentation variables with diagnostic testing results and clinical outcomes to derive guidelines that will enable safer and more efficient management of elderly patients presenting to the ED with abdominal pain and which is reproducible in the clinical setting in which it is used. Once we have developed and piloted the risk stratification project we will conduct a multi-center trial to measure the effectiveness of the risk stratification schema to achieve the stated objectives. The study will allow us to test the hypothesis that there are certain historical factors and physical examination findings that enable the physician to risk stratify senior patients presenting to the ED with abdominal pain thereby guiding the diagnostic work-up and improving health outcomes while optimizing resource use.
The pilot project that we are requesting funding for will develop and test a data-gathering tool for use in the larger Geriatric Abdominal Pain Risk Stratification (GAPRS) project. The APET pilot entails the development of a physician questionnaire whose intent is to capture information about usually undocumented factors that impact upon physician decision-making. The APET pilot data project expected completion date is January 2003. We plan to submit the full-scale grant application for the GAPRS risk stratification project in June of 2003. The agencies under consideration for this funding are the National Institute of Aging and the Hartford Foundation.
The pilot project entails the development of a physician questionnaire that will capture information about factors that are usually not well documented but reflect the physician’s decision-making process. Particularly, we are interested in what diagnostic tests are ordered, the reduction in diagnostic uncertainty due to imaging procedures, as well as decision-making about patient disposition The APET project will address the following hypotheses:
There are identifiable presentation variables that predict what diagnostic tests physicians order in the context of elderly patients presenting to the ED with abdominal pain.
The overall subjective physician impression for severity of illness at the time of initial history and physical influences decisions regarding diagnostic work-up.
Comorbidities, socio-economic factors, and patient preferences have influence on decisions regarding diagnostic work-up and disposition.
The suspected diagnosis and the degree of diagnostic certainty (following the initial history and physical) correlate with what diagnostic imaging tests are performed.
Diagnostic imaging studies decrease diagnostic uncertainty in elderly patients presenting to the ED with abdominal pain.
Once developed, the pilot questionnaire will be distributed on a patient by patient basis for a total sample size of 200 subjects. Inclusion criteria is age 60 years or older and a complaint of non-traumatic abdominal pain. Exclusion criteria include any abdominal surgical procedure within the past 30 days. All participants who meet inclusion criteria will be approached for consent to participate in the study. Eligible patients will initially be identified by chief complaint in the ED computer system and are consented by a research associate (RA). After initial physician examination of the patient the RA will have the physician complete the presentation and initial diagnostic test decision portions of the questionnaire. If the initial set of tests is inconclusive, the RA will have the physician complete the interim decision making section. The final section of the form, the disposition decision-making, is completed after the final diagnosis is made. Follow up phone calls will be made to participating subjects at 10 and 30 days to determine patient health status. The questionnaire results combined with chart review information will be recorded in a Microsoft Access database for statistical analysis looking at predictive and correlational measures.
Final Report:
Though there are several studies in the literature studies describing CT utilization in the ED, there is almost no data regarding which factors drive CT ordering or if CT improves diagnostic accuracy. To further evaluate clinical decision-making in the ED evaluation of abdominal pain, we developed a standardized method for obtaining real-time information regarding the physician’s clinical decision making. The form was developed, field tested, and modified before using it in the following study.
