Clement J. McDonald, MD
Grant Support: This work was supported by grant G08 LM008232 from the National Library of Medicine and grant 510040784 from the Indiana Twenty-First Century Research and Technology Fund. Funding for the Quality Grand Rounds series is supported by the California HealthCare Foundation as part of its Quality Initiative.
Potential Financial Conflicts of Interest: None disclosed.
Requests for Single Reprints: Clement J. McDonald, MD, Regenstrief Institute, 1050 Wishard Boulevard, Indianapolis, IN 46202; e-mail, firstname.lastname@example.org.
Increasing numbers of hospitals are implementing bar-coding systems to prevent errors in patient identification. In the present case, a diabetic patient admitted to a teaching hospital was mistakenly given the bar-coded identification wristband of another patient who was admitted at the same time. When a laboratory result that documented the diabetic patient's severe hyperglycemia was entered into the other patient's electronic medical record, the latter patient seemed to have a very high glucose level and was almost given what could have been a fatal dose of insulin. This near miss shows that computer systems, although having the potential to improve safety, may create new kinds of errors if not accompanied by well-designed, well-implemented cross-check processes and a culture of safety. Moreover, computer systems may have the pernicious effect of weakening human vigilance, removing an important safety protection. Researchers should continue to study real-world implementation of computerized systems to understand their benefits and potential harms, and administrators and providers should seek ways to anticipate these harms and mitigate them.
A typical bar code font (bottom), compared with the font size of a typical party name tag (top).
What really bothered me about the situation is that it just opened up so many avenues for mistakes, and the mistakes could be really serious medically. For instance, if we had taken that blood sugar at face value, we would have ordered a sliding-scale insulin regimen that would have given this patient over 10 units of insulin, which in a patient of normal blood sugar could have killed him because the blood sugar would drop so low that he could have gone into a coma.
The existing policy addresses this [2 patient identifiers requirement] in that the 2 forms of identification [required] before the original wristbands are placed is the critical step here and does need to be reiterated to the people at the frontlines who are placing the wristbands when patients are actually next to each other. So this is the critical first step and the downstream issues all relate to educational reminders and training for the staff, as opposed to any changes to the bar codes themselves or in identifiers that are available on the wristband.
In general, the system is really a good thing. I think it has dramatically decreased medication errors and it has dramatically decreased patients missing medication doses. [Although] it's not easy to learn and there are a lot of twists in the system that make it difficult for staff nurses … and it definitely increases their workload and takes more time away from direct patient care, it's a cost worth paying because of the increased safety to the patient … The first thing I recommend is don't install a faulty system, because what will happen is people will figure out a way to work around it. So if your system doesn't work right, or if it's too labor intensive, the staff will find a way to work around.
The case reinforced the idea that when labs, or some other objective evidence you get from a computer system doesn't quite match with what you are faced with clinically, you as a clinician have to step back and find out how you can put information together. Ask yourself, “Are these not true values?”, or is something just not right, and do things need to be rechecked … Even in a system that is supposed to work better than previous systems, there are still loopholes, and still things that need to be double-checked.
When people discover a discrepancy, it seems to be human nature to believe that there is a problem outside, that the person is wrong, if you will, that it can't possibly be the computer system, but something else. So I think there is, at times, a blind trust that the scanning system must be more accurate than the humans trying to rethink the process. And that's a very interesting phenomenon.
Edward C Wu
NYU School of Medicine
April 25, 2006
Corporate Strategies for Computerization
Corporate Strategies for Computerization
The critique of computerization by McDonald (1) highlights advice that all physicians would do well to heed. The ease of bar-coding can allow us to become automatons, shedding our critical thinking in the process. While McDonald touched upon technological alternatives and systematic changes that would improve such innovations as bar-coding, we suggest some additional lessons that medicine can learn from the corporate world.
Corporate principles can enhance patient safety while maintaining a culture of efficiency and effectiveness. The emergence of Six Sigma principles (2) and quality management departments as applied to medicine is promising. Kiosks have been used extensively at airports and financial institutions. The advent of a patient-centric kiosk, where patients self- input their information, has potentially large implications for efficiency and patient safety (3). Another common retail practice involves the repetition of an order. With the exception of blood product administration, this repetition is seldom used in medicine. To review a process with a patient sounds so simple, yet so difficult to enact in clinical practice. Medicine could learn from business by involving its customers "“ patients "“ more.
In addition to the technologies mentioned by McDonald, one that deserves mention is natural language processing (NLP). Still in its infancy, NLP can enhance adverse event detection (4). This technology utilizes computer algorithms to detect potential adverse events within the confines of an electronic medical record. The key challenge will be to integrate NLP into processes which inform providers in a timely and appropriate fashion, so as to prevent adverse events. While bar-coding can dull our sensibilities, NLP can heighten our senses to adverse events.
It is inevitable that new technologies will emerge as we shuttle patients through a hospital. How we implement these new technologies should not rest solely with a consultant, a vendor, or a committee. Our technology solutions should be looked upon in the same way that we care for patients"”from head to toe.
1. McDonald CJ. Computerization Can Create Safety Hazards: A Bar- Coding Near Miss. Ann Intern Med. 2006; 144:510-516. [PMID: 16585665]
2. Hagland M. Six Sigma: It's real, it's data-driven, and it's here. Health Care Strategic Management. 2005; 23: 12-16. [PMID: 16445106]
3. Porter SC, Cai Z, Gribbons W, Goldmann DA, Kohane IS. The asthma kiosk: a patient-centered technology for collaborative decision support in the emergency department. Journal of the American Medical Informatics Association. 2004;11:458-67. [PMID: 15298999]
4. Melton GB, Hripcsak G. Automated Detection of Adverse Events Using Natural Language Processing of Discharge Summaries. J Am Med Inform Assoc. 2005;12:448"“457. [PMID: 15802475]
Clement J McDonald
June 1, 2006
We thank Drs. Wu and Shah for their thoughtful letter. Just to be clear, I was not criticizing bar-code systems in general. Bar-coding is the perfect solution for many identification problems, e.g. the identification of groceries at the checkout counter and we all appreciate the resultant faster checkout times. My criticism was limited to bar coding systems used to verify medication dispensing at the bedside. (1) Both published reports and recent conversations with disappointed CEOs suggest that these systems have not delivered the nursing time savings, and error elimination they promised. Further, they appear to change nursing priorities - an un-intended consequence. I am sure that with time and improved technology these systems will meet their mark, but the available evidence contradicts the frothy hype for rushing to these systems today.
There is a general lesson here. The health care industry expects information technology (IT) "“ such as bar-coded medication dispensing and physician order entry "“ to solve many/most of its problems. These are plausible expectations and many will eventually be realized. However, today, the hype and the hope have run far ahead of the evidence and experience. We need to critically examine the benefits and harms of these systems as implemented in operating environments to determine what works and what doesn't and to identify design flaws and miss-assumptions as Patterson (2) did for bar coding and others have done for CPOE. Only with such knowledge will these products improve and reach their full promise.
The expectations about Six Sigma for health care may also be a bit frothy. Even in manufacturing the real achievement may only be four and a half, rather than Six, Sigma and some question its applicability to fields outside of manufacturing. (3) Given that health outcomes are not influenced by many fold increases in care process investment (4) and the fact that care costs have reached crisis levels - one might wonder whether this is the time to eliminate costly but marginal processes rather than investing even more to perfect them.
I agree that methods for capturing data derived from patients have value and am a fan of the work by Hripcsak, et al. (5) Natural language processing should become a major asset to medical information management.
1. McDonald CJ. Computerization Can Create Safety Hazards: A Bar- Coding Near Miss. Ann Intern Med. 2006; 144:510-516.
2. Patterson ES, Cook RI, Render ML. Improving patient safety by identifying side effects from introducing bar coding in medication administration. J AmMed Inform Assoc. 2002;9:540-53.
3. White E. Rethinking Quality Improvement. Wall Street Journal Sept. 19, 2005, page B3.
4. Sirovich BE, Gottlieb DJ, Welch HG, Fisher ES. Regional Variations in Health Care Intensity and Physician Perceptions of Quality of Care. Ann Int Med, May 2, 2006; Vol 144 (9); 641-649.
5. Hripcsak G, Austin JH, Alderson PO, Friedman C. Use of natural language processing to translate clinical information from a database of 889,921 chest radiographic reports. Radiology. 2002 Jul;224(1):157-63.
McDonald CJ. Computerization Can Create Safety Hazards: A Bar-Coding Near Miss. Ann Intern Med. 2006;144:510–516. doi: https://doi.org/10.7326/0003-4819-144-7-200604040-00010
Download citation file:
Published: Ann Intern Med. 2006;144(7):510-516.
Cardiology, Coronary Risk Factors, Diabetes, Endocrine and Metabolism, Hospital Medicine.
Results provided by:
Copyright © 2020 American College of Physicians. All Rights Reserved.
Print ISSN: 0003-4819 | Online ISSN: 1539-3704
Conditions of Use