It is time to shed some light on and refute some of the claims about the power of bar codes to deliver gains in patient safety. And, at the same time, the value of Lean and Six Sigma as significant tools in improving processes that actually can ensure patient safety needs to be reinforced.
The Food and Drug Administration (FDA) 2004 final ruling titled “Bar Code Label Requirements for Human Drug Products and Biological Products” is being touted as a necessary step for safety in the administration of patient care in the United States. The FDA “estimates that the bar code rule, once implemented, will result in more than 500,000 fewer adverse events during the next 20 years. Thus, FDA estimates a 50 percent reduction in medication errors…” FDA bar code compliance is required in spring 2006.
There is no question that using bar codes can have a positive effect on patient safety. Unfortunately, some suppliers and supporters of bar code systems seem to be exaggerating the role of bar codes. Merely adding a bar code system will not solve the problems of patient safety. The fact is that bar coding does not guarantee patient safety, it only guarantees consistent data capture. Simply put, a bar code will automate the process of recording static information.
Many hardware providers make claims of product safety and efficiency upon the mere application of a bar code solution. It is as if one has to simply introduce bar coding to their operations to realize the benefits. This is naïve at best and a gross misrepresentation at worst. The addition of the “patient safety” label also allows some providers to rationalize the sale of obsolete technology at premium prices.
The ability of a bar code to successfully capture and transmit data is dependent on:
Symbology: Bar codes use symbols to represent information. But the symbols can vary among countries, which potentially can generate misleading or incorrect data, negating the very advantages of bar coding in the first place.
Technology: The technology for bar coding is relatively simple and matured. Bar coding has been successfully applied in conventional manufacturing industries since the mid-1980s. A bar code reader can be obtained as an inexpensive peripheral to a conventional or portable computer system for a few hundred dollars. Software to translate the bar codes into legible text can be obtained as an open-source product or a commercial off-the-shelf product for a few hundred dollars per license.
It should be noted that a hospital or healthcare institute is usually better served by having their biomedical engineering and information technology groups collaborate to develop an internal solution. The hardware and software components and the development templates are readily available at a fraction of the cost charged by full-solution providers.
Readability: The information must be readable. This means that it must be protected from damage or deterioration. Hospitals frequently use sterilization or disinfectant solutions which can damage both the bar code reader and the label being read. Angular or cylinder variations in the labels can affect readability (very important to consider if the label is on a test tube or pill container). An unreadable barcode serves no purpose and generates no benefit, but instead causes delays as the healthcare practitioners must revert to manual methods.
Business rules: The business rules are necessary to make the data meaningful. For example, it is important to ensure that there is no confusion about the date because different regions use different formats for how dates are indicated. Does the bar code correctly distinguish June 7 from July 6? Also, does the bar code correctly accommodate names with special characters (i.e., accents, apostrophes)? Is there a mechanism to support truncation or partial matches? Without rules explicitly established, the bar code cannot even guarantee the consistent and correct capture of data.
The FDA press release, as well as releases from bar code systems providers (with vested interests in the proliferation of the technology), gives the impression that the mere adoption of this technology will be sufficient to generate the claimed reduction in adverse incidents. This is a significant exaggeration. The key to any safety initiative is the adoption of consistent processes that continually assess the biological, chemical and physical risks of the substance or specimens, and take steps to mitigate those risks through appropriate controls. At best, bar coding can play a role in this process by ensuring consistent data capture at appropriate control points in the process.
So what are the appropriate control points? This can be determined using a variety of Lean and Six Sigma techniques. Lean and Six Sigma are used predominantly in manufacturing, but have been much less frequently applied in the healthcare industry.
In order to have a stable process that ensures patient safety, the following steps need to be performed:
Value stream mapping is a Lean technique where the processes are analyzed to determine the value points, constraints and areas of waste. After the redundant or nonessential steps are removed, the process can be suitably prepared for further application to improve patient safety.
Product and process hazard analysis is conducted to determine the items of greatest risk. Techniques such as a failure modes and effect analysis can highlight and quantify hazards and determine the anticipated effect on the patient or specimen involved.
After the facility has completed the process and hazard analysis, the suitable mitigations can be determined and assigned by their priority. One of many mitigations includes the correct and consistent capture of patient and specimen information. It is only at this point where bar coding actually becomes relevant. Bar coding, along with the other mitigations, is incorporated into a revised and re-engineered process.
If a process or sequence of activities has been modified to promote value, minimize waste, mitigate hazards and maximize efficiencies, then that process must be validated and controlled. This will ensure that the core purpose of the process is fulfilled and no new issues have been introduced. Process controls include the audit trail of records necessary to demonstrate proof that the process was correctly completed. While bar-coding can support process control, the data obtained must be applied and transferred to the correct database fields to be analyzed and archived for future reference.
Long-term benefits can only be realized through a deliberate improvement initiative, which is where the benefits of a Six Sigma program can be manifested. After defining and measuring performance (incidents or near-misses, in the case of patient safety), a Six Sigma team can analyze, set targets and make improvements. The improvement steps will probably require a revision of the process and training, as well as additional resources and visual indicators. Bar coding may be a part of the solution.
Instead of just having a hardware provider or systems distributor install a bar code system without first ensuring the suitability of the processes and hazard mitigations, a more sound approach is to have a process expert work alongside the healthcare practitioners to determine the ideal process (independent of technology) that will best preserve patient safety.
Then, working in collaboration with the biomedical engineering and information technology groups, construct an internal technical solution that may or may not include bar coding. Other modes of data capture include RFID (radio frequency identification), which is becoming increasingly reliable and cost effective. An internal solution is flexible and adaptable without requiring the costly regulatory requirements of commercially distributed medical software – costs which must be borne by the end customer.
Process expertise can be readily obtained. For example, the ideal blood transfusion process was defined in 2003 in the Serious Hazards of Transfusion (SHOT) program by Professor Mike Murphy, et al, at John Radcliffe Hospital at Oxford University.) This process defines the critical points where a neutral data capture technology can be introduced to offset the process and product risks associated with blood transfusion.
A professor of blood transfusion medicine at the university, Murphy said, “We’ve looked very carefully at the hospital’s transfusion process, which is very complex and involves many staff. We’ve broken this down into a series of simpler steps and designed the software around them. The new system now means that one nurse is prompted to carry out 14 checks, where it previously required two nurses following 27 stages making paper records. It’s safer, it’s less time consuming and we know that staff prefer it.”
This is a successful example of first determining the hazards of a medical procedure, then defining a process, and finally applying bar coding technology to fit the process. Since the process was defined first, bar coding is one of several data capture methods that could be applied to give a similar result. Alternatively, if the process steps or business rules were compromised or neglected, the best bar coding system in the world would not make the patients any safer. Therefore, the claims of FDA and product marketers are refuted.
Reasoned insight on the application of a bar coding system to a healthcare facility to address safety concerns makes it clear that bar coding itself will not improve patient safety unless it has the suitable and compatible technology, symbology, interfaces and business rules for correct operation. Bar coding is worthwhile and applicable only when incorporated after the necessary work has been done to analyze and mitigate hazards, and define a sound process with critical control points for the timely application of data capture.
A bar coding solution need not be expensive. As an alternative to turnkey commercial solutions costing in the hundreds of thousands of dollars per user or license (and requiring constant regulatory approval costs for commercial development and testing), hospitals should assign a cross-functional team including a medical process expert and representatives from the clinical practitioners, biomedical engineering and information technology. This team could construct and maintain a customized solution from open-source technologies that would be adaptable and flexible to the needs of the facility.
Ultimately, patient safety is best guaranteed by removing risks and hazards at their root cause. This is the essence of Six Sigma and the only way to achieve the target of 500,000 fewer adverse events during the next 20 years and a 50 percent reduction in medication errors. There is no “silver bullet” technical solution to achieve these goals. A carefully considered, continually updated sequence of analysis and regular improvements of medical processes is the best hope for better patient safety.