NUR 6310 Module 2 Current-State Analysis Paper: Medication Reconciliation Workflow Through a Sociotechnical Lens Example

Three Lists, No Owner: A Current-State Analysis of Admission Medication Reconciliation at a Somerset Hospital

Student Name

M.S. in Nursing, Nursing Informatics, Thomas Edison State University

NUR-6310: Nursing Informatics: Systems Life Cycle

Mentor Name

Month Day, Year

Three Lists, No Owner: A Current-State Analysis of Admission Medication Reconciliation at a Somerset Hospital

Why Analyze First

Every person, unit and number in this paper is invented. A 290-bed hospital in Somerset plans to turn on a medication reconciliation module in its electronic health record. The vendor's default design assumes that one clinician collects the home medication list, a physician reviews each medication at admission and the list carries forward to discharge. The hospital's pharmacy and nursing leaders suspected that its actual process was different, and its safety reports showed eighteen events in a year in which a home medication was omitted, duplicated or given at the wrong dose after admission. The informatics nurse leading the analysis phase argued that configuring the module to match the policy rather than the work would automate the problems.

How the Analysis Was Done

Over three weeks, the informatics nurse observed 20 admissions from the emergency department to medical units, at different hours, following each home medication list from the first question to the first inpatient dose. She interviewed emergency nurses, admitting hospitalists, pharmacy technicians, floor nurses and four patients, collected the forms and screens in use and reviewed the eighteen safety reports. She recorded what people did, not what they said they did, and compared the two.

The Sociotechnical Model

Sittig and Singh (2010) proposed that health information technology be studied as part of a complex adaptive system in which eight dimensions shape each other. Three are technical (hardware and software, clinical content and the screen through which people use it). Two concern people and how their work and communication flow. Three concern the setting: the organization's own policies and culture, rules and pressures from outside, and how anyone checks whether the system is doing its job. They argued that problems usually arise at the intersections between dimensions, so that a technical fix to one part can fail because of another.

The analysis showed that in 14 of the 20 admissions, the home medication list was collected more than once by different people, and the versions differed in 9 of them. Physicians reconciled from whichever version appeared first on their screen. Nobody was responsible for deciding which list was right.

How Systems Introduce Errors

Ash et al. (2004) described two kinds of error that patient care information systems introduce. The first arises in entering and retrieving information: interfaces designed for quiet, uninterrupted work are used in busy settings, and structured fields can overload users and fragment the clinical picture. The second arises in communication and coordination: systems can represent work as a tidy sequence when real work is flexible and overlapping, and they can treat communication as a transfer of data when clinicians also need conversation. The authors argued that designers must understand real work before designing.

Both kinds appeared in Somerset. Home medication screens divided each medication into many fields, so busy emergency nurses entered drug names in a free-text comment instead, which the reconciliation screen could not read. And the record assumed a sequence from collection to verification to reconciliation, when in practice the steps overlapped and were repeated.

A Warning From Order Entry

Koppel et al. (2005) examined a common commercial order entry system in a teaching hospital by watching residents work, talking with them in groups and one by one and surveying them. Their list ran to 22 kinds of medication error risk made more likely by the system, grouped as information errors, such as fragmented displays that hid a patient's other medications, and human-machine interface flaws, such as screens that showed pharmacy inventory doses that physicians mistook for recommended doses. Many of the risks were reported by large shares of the house staff surveyed, and the study showed that a system meant to reduce errors could introduce new ones when its design did not match how clinicians worked.

The study's lesson for Somerset was direct: the vendor's default reconciliation screen showed home medications on one tab and inpatient orders on another, the kind of fragmented display Koppel's study linked to errors.

Requirements

The analysis produced requirements for the configuration team, each tied to a finding. One shared home medication list across the emergency department and inpatient units. A required source field for each entry and a status of verified or unverified. A side-by-side view of home medications and inpatient orders on a single screen. A standard pick list for common over-the-counter products and supplements. Pharmacy technician coverage for evenings and weekends, so that a trained person verifies the list for most admissions. A revised policy naming the pharmacy technician or admitting nurse as owner of the verified list and the physician as owner of reconciliation decisions. And a monthly accuracy audit comparing the verified list with a gold-standard interview on a sample of patients.

Validating the Findings

Before writing requirements, the informatics nurse took her findings back to the people she had observed. In two short sessions, emergency nurses, hospitalists and pharmacy technicians reviewed the eight-dimension table and corrected two points: night pharmacy coverage had been cut the previous year, not never offered, and hospitalists sometimes did call patients' pharmacies, but only when a list looked wrong. Both corrections changed the requirements. The sessions also built support for the redesign among the people who would use it, because they recognized their own work in the analysis.

Limits

Twenty admissions over three weeks may not capture rare situations, and the observer's presence may have changed behavior. The analysis focused on admission and did not follow medications to discharge.

Conclusion

Observation showed that admission reconciliation at Somerset ran on three lists with no clear owner. The sociotechnical model located problems in every dimension, the account of system-induced errors explained why free text and assumed sequences caused trouble and the order entry study warned against fragmented displays. Requirements built from these findings give the configuration team a design that fits the work, along with the staffing and policy changes no screen can provide.

References

Ash, J. S., Berg, M., & Coiera, E. (2004). Some unintended consequences of information technology in health care: The nature of patient care information system-related errors. Journal of the American Medical Informatics Association, 11(2), 104-112. https://doi.org/10.1197/jamia.M1471

Koppel, R., Metlay, J. P., Cohen, A., Abaluck, B., Localio, A. R., Kimmel, S. E., & Strom, B. L. (2005). Role of computerized physician order entry systems in facilitating medication errors. JAMA, 293(10), 1197-1203. https://doi.org/10.1001/jama.293.10.1197

Sittig, D. F., & Singh, H. (2010). A new sociotechnical model for studying health information technology in complex adaptive healthcare systems. Quality and Safety in Health Care, 19(Suppl 3), i68-i74. https://doi.org/10.1136/qshc.2010.042085