5 Questions: Lowe on hurdles facing electronic medical records

- By Tracie White

Steve Fisch Henry Lowe

Henry Lowe and colleagues recently reviewed all studies examining the use of computerized physician order systems in hospital ICUs.

The change to electronic medical records is slowly becoming a reality across the country. In an effort to examine how best to move forward with these far-reaching changes within hospitals, Henry Lowe, MD, conducted a Stanford study that reviewed all the scientific literature that has examined one area of electronic record keeping: the change within the intensive care units of hospitals from the traditional handwritten physician order notes to computerized physician order entry — doctors using computers to order medications, diagnostic tests and procedures.

Lowe, senior author of the study and senior associate dean for information resources and technology, discussed the study, which was published online in the Journal of Intensive Care Medicine on Jan. 21, with Inside Stanford Medicine writer Tracie White. (Other study authors included first author David Maslove, MD, a graduate student in biomedical informatics, and Norman Rizk, MD, professor of pulmonary and critical care medicine.) 

Q: The federal government’s 2009 economic stimulus package invested $19 billion in health information technology. Are we going to see immediate benefits from this initiative?

Lowe: Implementing health information technology, per se, may not produce major benefits. Instead the opportunities accrue from the longer-term changes in clinical culture and workflow that health information technology can provide. Commitments to patient safety, enhanced quality and improved clinical outcomes are essential to realizing its promise.

As with any new technology, there is a learning curve that must be overcome. In the case of physician order entry, unanticipated consequences may arise and need to be addressed before benefits are realized. In some cases, this process has taken months, or even years, and has sometimes required a profound restructuring of clinical workflow, or modification of the system itself. 

The benefits are likely to become evident only after practitioners have had some time to familiarize themselves with the new tools, become proficient in their use and adapt their workflows accordingly.

Q: Why did you choose to focus on the implementation of computerized physician order entry in intensive care units rather than hospitals as a whole?

Lowe: The ICU is a high-cost, data-intensive environment focused on the care of critically ill patients. It is also an environment in which effective clinical workflow is particularly important. Since one of the greatest potential risks from physician order entry is a negative impact on workflow, we wanted to review the entire literature in this arena to create a synthesis of what the evidence says about health information technology in this environment.

Because morbidity and mortality are high in the ICU, and because the risk of serious medical error is enhanced, outcomes of interest occur frequently and quickly, increasing the likelihood that differences with physician order entry are found. The ICU also presents a very stringent test of the new system, since the complex and stressful environment demand that systems be flexible, quick, reliable and unobtrusive

The first study to reveal the possible harms associated with physician order entry in fact came from a pediatric ICU, where mortality increased after the system was implemented. Subsequent studies in similar environments have either measured no negative impact or reported improvement in patient outcomes. The message from these studies is that the quality of the implementation is at least as important as the technology itself.

Q: What do you see as the most significant challenges facing intensive care units that hope to implement computerized prescription ordering systems? What are the potential benefits?

Lowe: The greatest challenges lie in finding ways to integrate physician order entry into existing workflows without causing them to break down. It can also be difficult to anticipate and plan for the consequences of deployment prior to the system going live. Ideally, clinical staff would see reductions in medical errors, enhancements in patient care and an easing of their workload, all of which will provide an incentive to persist with the new system despite the disruptions to workflow that are likely to occur.

To address this challenge, it’s desirable to study how the changes may impact the workflow both positively and negatively before implementation. Advanced training of staff is essential to minimize the negative impact to workflow.

Situations in which you seem to get the best results are those in which people use the technology not as an end in itself, but as a way to achieve desired outcomes. It is essential that medical staff “buy in” to the changes and are involved at every stage along the way.

Q: Do you think the potential benefits outweigh the possible pitfalls? Why?

Lowe: If approached cautiously, implementation in the ICU promises a number of benefits. These include reduced medication errors; better synthesis of complex clinical data trends (with appropriate alerts) and enhanced clinical decision-support capability.

Realizing these benefits depends on the software, the clinician’s ability to use the system and effective workflow integration.

A variety of potential harms have been reported in the literature, mostly related to missed medication doses and failures to renew prescriptions. Some of the benefits demonstrated have included a reduction in prescribing errors, more rational use of antibiotics and improved adherence to blood transfusion guidelines.

The challenge will be to implement these changes in a way that accounts for the complexity of workflow, generates buy-in from users, lessens the cognitive load associated with order entry and avoids burdening clinicians with clerical tasks. At the local level, small gains along the way will be helpful in convincing clinical and administrative staff of the potential benefits.

Q: Has the changeover to physicians using computers to order prescriptions been particularly slow in intensive care units, and if so, why?

Lowe: Specific rates of adoption in ICUs are not well-studied, but in the United States about 30 percent of hospitals have now adopted physician order entry in at least some clinical areas. However in mid-2010, only 14 percent of all U.S. hospitals had achieved the expected 10 percent level required for “meaningful use.” ICUs may see a lag in implementing physician order for medications, as some drugs that are commonly prescribed are administered by continuous IV infusion and adjusted frequently. This complicates the process of initiating, revising and recording medication administration orders. There may also be a reluctance to transition all medication ordering to electronic methods, since in the ICU, emergencies frequently arise which necessitate the immediate availability of a select number of drugs. Delayed availability of urgently needed medications has been thought to contribute to some of the harms associated with the implementation.

About Stanford Medicine

Stanford Medicine is an integrated academic health system comprising the Stanford School of Medicine and adult and pediatric health care delivery systems. Together, they harness the full potential of biomedicine through collaborative research, education and clinical care for patients. For more information, please visit med.stanford.edu.

2023 ISSUE 3

Exploring ways AI is applied to health care