Problem: Garrison Children's Hospital is a 225-bed hospital. Its 77-bed neonatal intensive care unit (NICU) provides care to the most fragile patients, premature and critically ill neonates. The 28-bed pediatric intensive care unit (PICU) cares for critically ill children from birth to eighteen years of age. Patients in this unit include those with life-threatening conditions that are acquired (trauma, child abuse, burns, surgical complications, and so forth) or congenital (congenital heart defects, craniofacial malformations, genetic disorders, inborn errors of metabolism, and so forth). Garrison is part of Premier Health Care, an academic medical center complex located in the Southeast. Premier Health Care also includes an adult hospital, a psychiatric hospital, and a full spectrum of adult and pediatric outpatient clinics. Within the past six months or so, Premier has implemented an electronic clinical documentation system in its adult hospital. More recently the same clinical documentation system has Case 11: Concerns and Workarounds with a Clinical Documentation System 461 been implemented at Garrison in both pediatric medical and surgery units and intensive care units. Electronic scheduling is to be implemented next. The adult hospital drives the decisions for the pediatric hospital, a circumstance that led to the adult hospital's CPOE vendor being chosen as the documentation vendor for both hospitals.

A CPOE system was implemented at Garrison Children's Hospital several years prior to implementation of the electronic clinical documentation system, which began in 2007. Information Systems Challenge A pressing challenge facing Garrison Children's Hospital is that nurses are very concerned and dissatisfied with the new clinical documentation system. They have voiced concerns formally to several nurse managers, and one nurse went directly to the chief nursing officer (CNO) stating that the "flow sheets" on the new system are grossly inadequate and she fears using them could lead to patient safety issues. Lunchroom conversations among nurses tend to center on their having no clear understanding of why the organization is automating clinical documentation or what it hopes to achieve. Nurses in the NICU and PICU seem to be most vocal about their concerns. They claim there is inconsistency in what is being documented and lack of standardization of content. The computer workstations are located outside the patients' rooms, so nurses generally document their notes on paper and then enter the data at the end of the shift or when they have time. The system support team, consisting of nurses as well as technology specialists, began the workflow analysis, system installation, staff training, and go-live first with a small number of units in both the adult hospital and the children's hospital, beginning in January 2007.

The NICU and PICU did not implement the system until May and June 2007. System support personnel moved rapidly through each unit, working to train and to manage questions. The timeline for each unit implementation was based on the number of beds in the unit and the number of staff to be trained. No consideration was given to staff members' prior experience with computers and keyboarding skills or to complexity of documentation and existing work processes. Although there are similarities between the adult and pediatric settings, there are also many differences in terms of unit design, computer resources (hardware), level of computer literacy, information documented, and work processes, not to mention patient populations. Little time was spent evaluating or planning for these differences and completing a thorough workflow analysis. After the initial units went live less and less time was spent on training and addressing unit-specific needs, due to the demands placed upon training staff to stay on the timeline in preparation for the next system implementation involving electronic scheduling. The clinical documentation system was implemented to the great consternation and dissatisfaction of the end users (physicians, nurses, social workers, and so forth) at Garrison, yet the Premier clinicians are happy with it.

Many Garrison physicians and nurses initially refused to use the system, stating it was "unsafe," "added to workload," and was not intuitive. A decision to stop using the system and return to the paper 462 Health IT Leadership documentation process was not then and is not now an option. Physician "champions" were encouraged to work with those who were recalcitrant and nursing staff were encouraged to "stick it out," in hopes that system use would "get easier." As a result, with their concerns and complaints essentially forced underground, Garrison clinical staff developed workarounds, morale was negatively affected, and the expectation that everyone would eventually "get it" and adapt has not become a reality. Instead, staff are writing on a self-created paper system and then translating those notes to the computer system; physicians are unable to retrieve important, timely patient information; and the time team members spend trying to retrieve pertinent patient information has increased. There have been clear instances where patient safety has been affected due to the problems with the appropriate use of this system.

Discussion Questions 1. What is the major problem in this case? What factors seem to have contributed to the current situation?

2. The nurses at Garrison argue that pediatric hospitals and intensive care units, in particular, are different from adult hospitals and that these differences should be clearly addressed in the implementation of a new clinical documentation system. Do you agree with this argument? Why or why not? Give examples from the literature to support your views.

3. How might the workflow issues and concerns mentioned in this case been detected earlier?

4. Assume you part of the leadership team at Garrison. How would you assess the current situation? What would you do first? Next? Explain what steps you would take and why you feel your approach is necessary.

5. What lessons can be learned from this case and applied to other settings?