Does inappropriate selectivity in information use relate to diagnostic errors and patient harm? The diagnosis of patients with dyspnea
Introduction
The diagnostic reasoning process is a complex process that involves many different decision making skills. Since diagnostic errors are often considered to be preventable and severe, studying the diagnostic process to find ways to reduce diagnostic error is important (Baker et al., 2004; Leape et al., 1991; Zwaan et al., 2010). Many different factors are involved in the occurrence of diagnostic errors, such as lack of knowledge or atypical presentation of the disease (Kostopoulou, Mousoulis, & Delaney, 2009; Neale, Woloshynowych, & Vincent, 2001; Schiff et al., 2009; Zwaan et al., 2010). In addition, research shows that in many cases in which a diagnostic error occurred, the physician did not consider the correct diagnosis from the start, which is often caused by cognitive biases (Berner & Graber, 2008; Croskerry, 2003; Elstein, 1999). Cognitive biases are faulty beliefs that affect decision making and occur because physicians use heuristics during the diagnostic process (Bornstein & Emler, 2001; Elstein, 1999; Tversky & Kahneman, 1974). Heuristics are shortcuts in the reasoning process, which means that not all available information is gathered or used to come to a diagnosis (Wegwarth, Gaissmaier, & Gigenrenzer, 2009). These heuristics are necessary to diagnose a patient within a reasonable amount of time and without conducting many unnecessary diagnostic tests. Heuristics are usually associated with fast diagnostic reasoning, and in most cases with correct diagnoses (Orient, 2009). Actually, the use of heuristics in the diagnostic process, can even lead to better diagnoses (Wegwarth et al., 2009). Particularly experts are able to diagnose a patient after gathering little data because they rely on the heuristic of pattern recognition (Groves, O'Rourke, & Alexander, 2003).
Although the use of heuristics is important in diagnostic reasoning, they may lead to faulty data-gathering, faulty data-synthesis and diagnostic errors (Graber, Franklin, & Gordon, 2005). For example, when the physician focuses on a specific diagnosis based on his/her recent experiences, and as a consequence does not adequately evaluate the evidence pointing towards alternatives (availability bias) (Berner & Graber, 2008; Berner, Maisiak, Heuderbert, & Young, Jr., 2003; Graber, 2005). Research showed that many cognitive biases occur in diagnostic reasoning and that they occur at all levels of expertise (Dubeau, Voytovich, & Rippey, 1986; Graber, Gordon, & Franklin, 2002; Redelmeier, 2005; Voytovich, Rippey, & Suffredini, 1985). The common denominator of most of the cognitive biases is that physicians are too selective in their reasoning process and therefore overlook likely diagnoses. This selectivity is inappropriate, which may result in a diagnostic error when relevant information is missed (Elstein, 1999). Inappropriate selectivity may have serious consequences depending on the stage of the diagnostic process in which it occurs and strategies to prevent inappropriate selectivity should be adapted to the specific stage (Croskerry, 2003; Kempainen, Migeon, & Wolf, 2003). Inappropriate selectivity in clinical practice and the situations in which it leads to diagnostic error and patient harm has not been studied extensively (Kostopoulou et al., 2009).
In our previously published study, we found that physicians often gathered insufficient information or lacked to follow-up on relevant findings (Zwaan, Thijs, Wagner, Van der Wal, & Timmermans, 2012). When we asked the physicians about these suboptimal decisions, we learned that most of these were deliberate decisions. However, we did not examine whether this was either due to a lack of knowledge or whether the physicians had been too selective throughout the process. Therefore, for this study we conducted further analysis on the data to determine: 1. the occurrence of inappropriate selectivity in the information-gathering and information-processing stages of the diagnostic reasoning process of dyspnea patients, and 2. to what extent inappropriate selectivity in the diagnostic reasoning process is related to diagnostic error and patient harm in clinical practice.
Section snippets
Method
Patient record reviews in combination with interviews with the treating physicians were used to determine the occurrence of inappropriate selectivity, the stages of the process in which selectivity occurred and the occurrence of diagnostic error and patient harm. See Fig. 1 for an overview of the data-gathering process of the study.
Patient sample
The patient characteristics are described in Table 1.
Physicians
The physicians who included the patients in the study and who were interviewed involved 72 medical residents who were supervised by a medical specialist. The residents had on average 29 months (SD = 26.5) of work experience.
Inappropriate selective diagnostic reasoning
In 45.7% (113 of 247 cases) inappropriate selectivity occurred. This involved selective information-gathering in 33.2% (82 of 247) of the cases while selective information-processing was identified in 12.6% (31 of 247) of
Discussion
Inappropriate selectivity occurred frequently in the diagnostic reasoning process and was associated with the occurrence of more diagnostic errors and patient harm compared to cases without it. Inappropriate selective information-gathering occurred more often than inappropriate selective information-processing. However adverse outcomes were more often associated with selective information-processing than with selective information-gathering.
There are some limitations of this study. First,
Conclusions
Inappropriate selectivity in the diagnostic reasoning process occurs frequently and is related to diagnostic error and patient harm. This is especially true for inappropriate selectivity in the information-processing stages of the diagnostic process. In order to improve diagnostic reasoning and reduce the occurrence of diagnostic errors and patient harm systematic feedback during supervision and handovers could be helpful. Furthermore, more falsifying reasoning strategies may help to detect a
Acknowledgments
The authors would like to thank: S. Lubberding, MSc for her help with the data gathering. Y. Smulders, MD, PhD for his comments on an earlier version of the manuscript. The physicians who reviewed the patient records: H.A.P Asscheman, MD, PhD; W.W. Meijer, MD; K.J. Roozendaal, MD, PhD and J. Silberbusch, MD, PhD. We would also like to thank the hospitals and their staff for including the patients in the study and for facilitating the patient records.
The study on which the manuscript is based
References (30)
- et al.
Overconfidence as a cause of diagnostic error in medicine
American Journal of Medicine
(2008) Diagnostic errors in medicine: a case of neglect
The Joint Commission Journal on Quality and Patient Safety
(2005)- et al.
A theoretical analysis of insight into a reasoning task
Cognitive Psychology
(1970) - et al.
The distortion of information during decisions
Organizational Bahavior and Human Decision Processes
(1996) - et al.
The Canadian Adverse Events Study: the incidence of adverse events among hospital patients in Canada
Canadian Medical Association Journal
(2004) - et al.
Clinician performance and prominence of diagnoses displayed by a clinical diagnostic decision support system
AMIA Annual Symposium Proceedings
(2003) - et al.
Rationality in medical decision making: a review of the literature on doctors' decision-making biases
Journal of Evaluation in Clinical Practice
(2001) The importance of cognitive errors in diagnosis and strategies to minimize them
Academic Medicine
(2003)- et al.
Premature conclusions in the diagnosis of iron-deficiency anemia: cause and effect
Medical Decision Making
(1986) Heuristics and Biases: selected errors in clinical reasoning
Academic Medicine
(1999)