Decision Support: Sure-Footed Steps toward Clinical Process Improvement

May / June 2009
Decision Support

Sure-Footed Steps toward Clinical Process Improvement

The information technology (IT) plan in most hospitals is beginning to look like the wish list of your average American teenager — software, PDAs, wireless infrastructure. Healthcare is amidst a revolutionary game of catch-up when it comes to applying information technology to the business of patient care. Today’s IT plans reflect ambitious, enterprise-wide adoption of electronic medical records (EMR) and their supporting pillars: physician and nursing documentation, electronic reminder systems, computerized prescriber order-entry (CPOE) and test-result reporting (Keyhani et al., 2008). However, these ambitions are tempered by capital constraints and the desire for short-term returns on investment.

Current trends in payment models and performance transparency place unprecedented emphasis on the accuracy and accessibility of comprehensive medical data in hospitals. As such, modernizing the electronic healthcare infrastructure has become a presidential decree to mitigate run-away healthcare costs and stimulate a lagging economy. Unfortunately, industry experts have responded with measured enthusiasm. A January report from the National Academies of Science (NAS) suggested that U.S. hospitals are not ready to pursue rapid EMR deployment given currently available technologies (Stead & Lin, 2009).

NAS reported that current healthcare IT solutions by and large focus on recreating paper-based system as electronic documents without improving information utility. Furthermore, the researchers conclude that technology has been driven more by business objectives, such as mandatory reporting, than by a need for clinical care improvement. The trickle-down effect has been that caregivers are expected to painstakingly document online merely to comply with regulations or to defend against lawsuits, not to produce higher quality healthcare and enhanced care processes.

As a result, many current systems pack unfortunate unintended consequences. Without appreciable clinical, financial, or workflow incentives, caregivers are likely to circumvent technology. The ramifications of system workarounds and poor user-centric design have been well documented. The Veteran’s Health Administration faulted design flaws in their early barcode medication administration system for non-compliance and workarounds Patterson et al., 2002). The Leapfrog Group has cited CPOE system design flaw for low adoption rates and physician reluctance (2008). In one case, researchers from the University of Pittsburgh School of Medicine correlated CPOE use with an increase in pediatric patient mortality citing disruptions in caregiver communication and delays in treatment (Han et al., 2005).

To guard against user revolt and circumvention, IT must deliver value to the caregiver. Missing from the EMR equation is a focus on cognitive support for caregivers when and where they need it — from simple performance aids to more complex configurable display of evidence-based protocols. The delivery of sound decision support has been constrained by two major factors: infrastructure challenges in serving the diverse needs of a mobile workforce and the strenuous intellectual investment required to establish a hospital’s cognitive support knowledgebase. Neither prerequisite is easily accomplished. Providing electronic cognitive support to caregivers first requires research and consensus-building across the medical disciplines — no simple matter. A 2002 study identified 59 obstacles to defining a single evidence-based protocol (Ely et al.). Compounded with human nature and organizational politics, the process can be formidable. Yet, the exercise is absolutely critical. For example, a drug library of safe dose ranges or a standardized diagnosis-related order set build on clinical best practice is a highly valuable cognitive support tool regardless of an organizations technology adoption. However, the development of these cognitive support knowledge bases generally coincides with and complicates the adoption of technology because they are essential for CPOE to operate.

Fortunately, putting cognitive support into the hands of clinicians does not require enterprise-wide EMR adoption. Immediate and impactful progress can be set in motion through smaller steps leveraging “elementary technologies. “Several tools currently exist in clinical practice that exhibit the “triple-threat” characteristics of an elementary technology: they address a discrete need by delivering hospital-defined cognitive support, meet with a high degree of user acceptance, and require minimal IT investment and administration. Two emergent technologies stand out as compelling examples of sure-footed steps toward better delivery of cognitive support.

Pharmaceutical Algorithm Computerized Calculations
The average 200-bed hospital administers over one million medication doses annually. An estimated 10 to 20% of these doses are calculated at the point of care by nurses using dosing charts, standard calculators, or manual computation. It is approximated that miscalculations lead to patient harm in 2.25 of every 1,000 doses given but these data are likely under-reported since lack of documentation renders these errors undiscoverable. Even with flawless math skills, nurses must further verify that the physician’s ordered dose is appropriate and recall any medication-specific considerations such as the correct route of administration. In spite of the many technologies that safeguard the medication use process — CPOE, barcode point of care systems, and nursing documentation — the nurse remains without adequate cognitive support during dose volume calculation — an especially risky, unnerving necessity in the treatment of intensive care and pediatric patients.

Katharine Francis, RN, a neonatal nurse, recognized the risks associated with this routine nursing practice and created a unique dose calculator for nurses with embedded cognitive support including a hospital-defined drug library of safe dose ranges. The pharmaceutical algorithm computerized calculator (pac2) from InformMed, Inc. is deployed on a handheld device at the point of care to assist in injectible medication administrations and the verification of safe infusion pump programming. As a nurse calculates a dose volume per physician orders, the pac2 automatically performs unit-conversion equations while calculating the correct dose volume. The software intercepts common entry errors such as misplaced decimal points or improper expression factors, questions inappropriate drug orders, gives access to essential dosing information and provides notification when calculated doses fall outside of the established limits for the drug ordered. In addition, the system provides an audit trail of each calculation and dose volume administered as a by-product of use. This solution effectively provides peace of mind to nurses “at the sharp end” allowing them to focus on the care that only they can deliver to their patients.



patient safety and quality healthcare
InformMed’s pharmaceutical algorithm computerized calculator (pac2).
Photo Courtesy of InformMed

In clinical pediatric use within the Children’s Hospital of Illinois at OSF Saint Francis Medical Center, Peoria IL, the pac2 has demonstrated a 95% reduction in dose errors (Torres & Henricks, 2008). Nurses prefer to use the pac2 over traditional practices because it affords them assurance of safe doses without adding time to their administration process. InformMed reports that the standard implementation process requires fewer than four weeks, no system interfacing, and typically less than one hour of training per nurse. This quick installation is made possible because, as a stand-alone solution, there is no need for system interfacing. Additionally, nurses can be rapidly trained in one hour-long sessions because the pac2 user interface mirrors the manual calculation process.

Automated Admissions Decision Support
The majority of physician orders for a newly hospitalized patient are written at the time of admission. The initial care plan greatly influences the patient’s overall clinical outcome. Omitted tests, duplicative medication therapies, failure to set dietary guidelines, or delayed scheduling of discharge transfer orders represent some of the many opportunities for inefficient and unsafe processes to negatively impact the quality of care.

In all but a small number of the most wired hospitals, admission orders are still handwritten into a paper chart. To help comply with accepted evidence-based protocols for common diagnosis, such as community acquired pneumonia, physicians use paper-based forms to guide their orders. As most patients present with multiple diagnosis, the doctor must reconcile the protocols manually. When complete, admission orders can be a dozen or more pages of handwritten notes or manually collated forms, and represent more than 100 minutes of the physician’s day. Omissions, incomplete and illegible orders further add to the physician’s and other caregivers’ time burden over the course of the patient’s care. What’s more, problem orders may contribute to a serious adverse events and patient harm.

An automated admissions decision support (AADS) tool guides admitting physicians and hospitalists though the time-consuming process of placing admission orders. Comprising more than 180 diagnoses-related order sets, the Order Optimizer AADS system created by Intercede Health provides cognitive support for inpatient orders per hospital-configurable protocols. By merging the required orders for patients with multiple-diagnosis, alerting the prescriber to drug interactions, and prompting the physician to pro-actively schedule a transfer to skilled nursing facility, for example, the Order Optimizer demonstrates an ability to reduce order time by 67% and shorten inpatient stays (Correspondence with Order Optimizer, Inc. per time and motion study conducted by Dr. Aaron Rosenberg, lead hospitalist, Bon Secours Medical Center, Richmond Virginia, April 22, 2008).

Unlike full-featured CPOE systems, AADS is typical of elementary technology because it fits into the paper-based world, which is the status quo for the vast majority of hospitals. Completed orders are printed in a concise, legible format and placed in the medical record. Electronic copies can be shared online. Because of the documented time-savings for physicians, users embrace the tool while the hospital benefits from best-practice compliance. Like the InformMed solution, Order Optimizer is capable of operating without interfaces, though data exchange is support through HL7 messaging, and implementation guided by a comprehensive starter set of evidence-based protocols is a matter of weeks. Physician training is accomplished in less than one week, and the company’s application service provider (ASP) model alleviates IT support demands on hospital staff.

Commonalities of Success
The pac2 and Order Optimizer are just two examples of tools which focus on cognitive support. Both systems are effectively training clinical users for an increasingly intensive digital future. But unlike enterprise systems which can be intrusive to clinical workflow, these systems serve the end user by saving time, reducing errors, and delivering optimal care for the patient. Because they accomplish a specific, yet valuable objective, users learn to employ the systems to their full potential in minutes, with minimal drain on IT resources and capital budgets.

Moreover, hospitals using these elementary technologies are gradually accomplishing the intellectual heavy-lifting necessary for broader cognitive support applications, e.g., drug libraries for smart infusion and evidence-based protocols to fuel future CPOE expansion. When the time comes to implement smart pumps and CPOE, these hospitals will be prepared to move quickly and confidently.

As stand-alone systems in an age of integration, elementary technologies fill holes in core system vendor offerings. With prudent development and core system vendor cooperation, these tools will serve clinicians for many years to come.

Jamie Kelly is the president of Entropy Research, Inc, a marketing firm serving health care information technology vendors, and a long-time evangelist for patient safety improvement through prudent technology adoption. Kelly is also co-founder of the annual unSUMMIT for Bedside Barcoding educational conference ( She may be contacted at


Ely, J. W., Osheroff, J. SA., Ebell, M. H., Chambliss, M. L., Vinson, D. C., Stevermer, J. J., & Pifer, E. A. (2002). Obstacles to answering doctors’ questions about patient care with evidence: Qualitative study. BMJ, 324, 710. Available at

Han, Y. Y., Carcillo, J. A., Venkataraman, S. T., Clark, R. S., Watson, R. S., Nguyen, T. C., Bayir, H., & Orr, R. A. (2005, December). Unexpected increased mortality after implementation of a commercially sold computerized physician order entry system. Pediatrics, 116(6), 1506-1512.

Keyhani, S., Hebert, P. L., Ross, J. S., Federman, A., Zhu, C. W., Siu, A. L. (200*, December). Electronic health record components and the quality of care. Medical Care, 46(12), 1267 — 1272.

Leapfrog Issues a caution that quality assurance necessary during implementation of computerized medication systems. (2008, October 13). Washington, DC: The Leapfrog Group. Available at

Patterson, E. S., Cook, R. I., & Render, M. L. (2002, September/October). Improving patient safety by identifying side effects from introducing bar coding in medication administration. Journal of the American Medical Informatics Association, 9(5), 540-553.

Stead, W. W. & Lin, H. S. [Eds.]. (2009, January). Computational technology for effective health care: Immediate steps and strategic directions. Committee on Engaging the Computer Science Research Community in Health Care Informatics; National Research Council.

Torres, A. & Henricks, C. (2008). FMEA analysis of risk reduction during emergent IV medication administration. OSF Saint Francis Medical Center, Children’s Hospital of Illinois. Institute for Healthcare Improvement’s (IHI) 2008 Annual National Forum on Quality Improvement in Health Care.