HCLF: The human aspects of decision support

CHICAGO—Studying people in natural clinical settings help in the development of guidelines and optimal interventions, explained Vimla L. Patel, PhD, DSc, senior research scientist at the New York Academy of Medicine, speaking during the Healthcare Leadership Forum on Nov. 14.

For example, she performed collaborative research with the American College of Physicians as part of an effort to develop methods for encoding guidelines and incorporating them into EHRs and CPOE systems. “We observed in detail the creation of clinical guidelines in their Clinical Efficacy Assessment Program and studied their interpretation and use by different classes of physician users—experts and generalists.”

One myth is that clinicians are optimal decision makers, where their actions can be entirely specified through standardized procedures, protocols, rules and norms. Patel’s research, however, has shown that use of evidence varies across clinical settings. Epidemiologic evidence in probalistic form is used less easily and therefore less frequently than might be desired by those promoting more use of evidence-based medicine (EBM). “It requires more thought about how to make EBM more usable and accessible in busy clinical environments.”

Another myth is that deviations from evidence-based protocols are both a cause of lower performance and the main threat to safety.

A deviation from protocol is defined, she explained, as an error if:

  • there was a violation of a prescribed order of activities
  • the violation resulted in compromising patient care

For example, a resident attempts to remove the spine board underneath a patient before the patient’s spine was confirmed as not injured.

The researchers defined deviations as innovations when the deviation could potentially benefit the patient and brings a novel perspective to the situation. For example, when attempting to diagnose the cause for head injuries in a patient, the resident noticed that the 1-week-old head trauma did not look like a typical presentation. In addition, the patient had a high Glasgow Coma score, was lucid and conscious, but had a wound on the leg, and had presented with high temperature. The resident did not do the guideline-required head x-ray, instead requesting blood cultures for the patient. The problem resolved after blood culture result showed acute infection and appropriate antibiotics were instituted.

Clinicians deviate from routine or standard practice under conditions of uncertainty and ambiguity as in the trauma care setting, Patel said. “These deviations when made by experts were shown to result mostly in innovations. These deviations when made by trainees were shown to result mostly in errors.”

The art of medicine will be lost if all actions have to conform to a regularity of standards, she said. The use of guidelines is also culturally driven, and culture currently provides flexibility for data interpretation to construct meaning in any situation. Evidence-based guidelines support, but do not replace, the exploratory and creative nature of human problem solvers as clinicians acquire expertise in their domains, they generate fewer errors than trainees do. Experts in their own domain of expertise make as many errors as some of the trainees, but they detect the problem and recover very quickly from the errors, generally before damage can occur. The novices are less able to detect their own errors or to recover from them, Patel said.

“Error detection and recovery are part of development of expertise in medicine. Human beings have a capacity to recover from error and these abilities are an important contributor to safety and not just a source of error. This suggests the important role of training through simulations.”

In conclusion, Patel said that following evidence-based guidelines and standards minimizes errors and provides care consistency. “Deviation from standards when appropriate is critical in building flexibility and adaptiveness, which are characteristics of expertise. The right technological interventions require understanding the boundaries of safe practice for practitioners (human resiliency) before developing and introducing tools for human use. This argues for assuring that systems for decision support be similarly resilient. EBM is a method for decision support of, but not a replacement for, an inherently human set of problem-solving capabilities.”

The Healthcare Leadership Forum was sponsored by ClinicalKey and presented by Clinical Innovation + Technology.