© 2014 by the President and Fellows of Harvard College. This case was written by Lina Matta, PharmD, and Mark Bloomberg, MD, Harvard T.H. Chan School of Public Health, as the basis for class discussion and education rather than to illustrate either effective or ineffective handling of an administrative or public health situation. This publication may not be digitized, photocopied, or otherwise reproduced, posted, or transmitted, without the permission of Harvard T.H. Chan School of Public Health.

PH4-016 2014

Medication Management at Acme Medical Center “This patient might have died from complications related to her TPN infusion,” said Dr. Isaac Johnson, Chief Medical Officer at Acme Medical Center (AMC). His review of the root cause analysis of a recent adverse event in medication management at the children’s hospital within AMC suggested that his team’s efforts to get every employee engaged in quality improvement were inadequate to address the complexity of care provided at a tertiary facility like AMC. His partner in this effort, Chief Nursing Officer Sheryl Turner, noted, “There is a real need to hardwire into the institution how quality improvement is to be performed.” While the need was evident, it was not as clear what the next steps should be.

Acme Medical Center

Acme Medical Center (AMC) was located in Middle City and enjoyed a long and proud history. The second oldest hospital in the state, it was founded in 1876 as The City Clinic with the support of local business and society leaders as initial sponsors. Since then, Acme Medical Center grew to become a 515-bed tertiary center that provided a complete range of medical services to adults. Children were also cared for through its adjacent 100-bed pediatric institution, The Acme Hospital for Children. Acme Medical Center was organized much like most other academic institutions. While independently governed, it maintained close ties with its next-door neighbor Acme University School of Medicine. Its salaried clinical staff was organized into a separate faculty practice known as the Acme Medical Center Physicians Organization. The President of the faculty practice reported directly to the CEO of AMC and was equal in position to AMC’s COO, CNO, and CFO. In addition to its tertiary services, Acme Medical Center had a well-deserved reputation for serving as the local hospital for one of the city’s ethnic communities whose urban enclave surrounded the hospital complex. This created a culture within which AMC staff viewed themselves as more than a purely academic medical center, taking great pride in fulfilling this additional local mission. While research was fostered and conducted on campus, the staff of AMC appreciated having this local primary care mission and proud of being less formal and more relaxed than some its academic neighbors. Most of the staff, both non-clinical and clinical, had a long history with the institution and were strong supporters of its mission and the care provided to patients.

This document is authorized for use only in Kris Michaelson’s HCM520-WI17D course at Colorado State University – Global Campus, from January 2018 to July 2018.

Medication Management at Acme Medical Center PH4-016

2

Quality Improvement at Acme Medical Center

David Herman, CEO of Acme Medical Center (AMC), set the tone for how he hoped quality improvement would be addressed within the hospital. His view was that, “We can’t run from our mistakes. We need to be fair and consistent in our approach and need to give our staff a safe place to recognize and respond to medical errors.” The hospital had a patient care assessment committee, whose duties included collecting the details of all significant adverse events, notifying the state medical board of their occurrence, and then following up on any queries that ensued. To date, quality improvement was largely left to the initiative of each department. The long-time former Director of QA was particularly focused on meeting the standards of the joint commission. She believed that quality improvement efforts should be left to the discretion of the front-line departments who best understood their own processes. Dr. Isaac Johnson had been the Chief Medical Officer of Acme Medical Center for the last three years. In that capacity he had responsibility for both patient safety and ongoing quality improvement activities. Sheryl Turner had been the Chief Nursing Officer for the last five years and was responsible for all of the patient care services throughout the institution. This included all medical and surgical units, all ancillary services, and all ambulatory services. The CNO/CMO partnership formed between Turner and Johnson had become a close one. Two years ago they initiated efforts to educate and inform their fellow senior managers about the impact that quality issues have on the institution and the need to create cultural change to have a decisive impact on safety and quality. While their fellow executives voiced support, the cultural change they desired had not yet materialized. To that end, last year they established a new Department of Quality and Patient Safety and replaced the former Director of QA with a new person with experience in long-term quality improvement. Since then, AMC had established “charters” which were formal statements of authorization for addressing specific institution-wide quality improvement initiatives. The first of these was related to medication errors and resulted in better efforts to identify all errors and evaluate the root causes that contributed to their occurrence. The charter outlined the system- wide educational programs, policies, and procedural changes that were designed to elevate awareness of the issue and promote ways to identify all such errors. Both CNO Turner and CMO Johnson also wanted to extend deeper awareness of QI and patient safety beyond their fellow senior managers and to build a greater commitment at the board level. While a few initial steps were taken to do this, such as reporting error rates and discussing the most common and the most serious errors, almost all of the board involvement in QI remained based in the board’s quality committee and did not extend to the larger board. The full board only received a summary report on quality every two months. Ms. Turner and Dr. Johnson often discussed how best to develop the board as a motivating force in QI and patient safety, but they perceived that the board was far more comfortable discussing financial issues and strategic direction. Most laypersons on the board lacked a working understanding of QI and

This document is authorized for use only in Kris Michaelson’s HCM520-WI17D course at Colorado State University – Global Campus, from January 2018 to July 2018.

Medication Management at Acme Medical Center PH4-016

3

patient safety issues, and some felt that quality issues should be left to the clinical staff of the hospital to address.

Medication Management and Its Role at Acme Medical Center

As in other medical institutions, medication management at AMC was considered to be an essential component of good medical care. Several important aspects of medication management were actively used in AMC with the goal of optimizing therapeutic regimens that would improve health outcomes. These included the medication use process (MUP) (see Exhibit 1), evaluation of a patient’s medication regimen, and documentation and follow-up of therapeutic progress and adverse events. The MUP represented a series of steps that began with an order for a medication and ultimately resulted in the administration of a medication to the patient. According to the Institute of Medicine, although errors within the medication use process can occur at any stage, they occurred most frequently during the prescribing and administering stage. The details of the MUP differed between pediatric and adult patients and also between orders for intravenous medications, total parenteral nutrition (TPN), and oral medications. The MUP process for pediatric TPN is detailed in Exhibit 2. Ralph Rolls, Director of Pharmacy at AMC, described the pediatric pharmacy staffing at AMC. On a weekday day shift, there were two or three pediatric pharmacists on the patient floor with at least one of them actually stationed on the floor, providing easy access to the team. On a weekday evening shift, there was one pediatric pharmacist available in central pharmacy and no pharmacist on the patient floors. On the weekends, there was one pediatric pharmacist assigned for eight hours out of every 24 hours. This pharmacist may be on the day or evening shift but not the night shift. On a weekday night shift, the entire hospital (adult and pediatric) was supported by two general pharmacists. On the weekend night shift, the hospital was supported by one general pharmacist. On weekdays the dietician, who was the expert in TPN formulation, was available to assist medical residents in preparing TPN orders; the resident then wrote the actual order. On weekends, the dietician was on-call but not on-site. The number of TPN orders did not vary by day of the week. All TPN orders had to be transmitted to the outside vendor by 1 p.m., seven days a week.

The Adverse Event

Susan was a vivacious 16-year old who was diagnosed with non-specific colitis several years before the event. She had been admitted to the inpatient service a week earlier and then discharged. Once home she began to have increasing pain and required readmission on a Friday, when efforts to place a peripherally-inserted central line failed. On Saturday, Susan was taken to the OR for placement of a Broviac catheter so she could be maintained on total parenteral nutrition. The line was placed without incident and TPN was begun. Several hours later, the nurse noticed a precipitate in the tubing and a tendency for the line to clog. She changed the filter but, despite multiple attempts to clear it, the central line clogged again and became

This document is authorized for use only in Kris Michaelson’s HCM520-WI17D course at Colorado State University – Global Campus, from January 2018 to July 2018.

Medication Management at Acme Medical Center PH4-016

4

inoperative. Susan had to be returned to the operating room for removal of the clogged line and the insertion of a new one. When the removed clogged line was examined, it was found to be plugged with particulate matter that had apparently precipitated out of the TPN solution. In addition, the partially full TPN bag itself was placed in the medication room, and by the next morning a precipitate had separated and settled into the bottom of the bag. When this was evaluated, it was discovered that the calcium and phosphate in the TPN solution were outside of established parameters and had created the precipitate that caused the line to clog. Susan recovered uneventfully and tolerated her TPN without further incident. The clinical case was routinely entered into the AIMS (Adverse Incident Monitoring System) software the hospital used to track such events and was flagged by risk management for evaluation.

The Root Cause Analysis Process

One of the important processes used to follow up on any adverse events at AMC was an evaluation by the risk management staff, during which a decision was made on whether to undertake a root cause analysis (RCA). An RCA involved investigating the potential causes of an undesirable outcome, including those related to medication errors (see Exhibit 3). Adverse events that resulted in significant impact to a patient and/or seemed to be a repetition of a lesser event were normally subjected to root cause analysis to determine the reason(s) the event occurred. The results of the RCA were used to redesign care processes, practices, and policies in order to prevent future similar adverse events. The immediate goal of an RCA was to focus on systems and processes without emphasis on the mistakes of an individual. The long-term goal was to eliminate the “root cause” of the problem or error. The process required a comprehensive and team-based approach to the problem that systematically investigated each step of the process. In the past two years, over 50 RCAs had been performed, and these had resulted in numerous procedural changes intended to reduce the likelihood of similar errors. Nonetheless, the occurrence of errors continued, and the pace of approximately one RCA every two weeks had continued unchanged. At AMC, all relevant team members were included in the initial RCA and given a task list to complete. A successful RCA should go beyond the initial meeting, including a follow-up process where each participant reported findings and outcomes of changes implemented. However, one of the participants at a recent AMC RCA expressed frustration at the lack of timely and comprehensive follow-up, describing the process as “untied loose ends.” The RCA performed following the TPN incident noted the following:

1. On the Friday of admission, the initial development of the TPN recipe was done by a dietician, and a pediatric resident completed the order, but because the initial line placement was unsuccessful, that order was not filled.

2. The next day, the TPN was re-ordered once the Broviac catheter had been placed, but because the patient’s calcium had dropped, the pediatric resident changed the order to

This document is authorized for use only in Kris Michaelson’s HCM520-WI17D course at Colorado State University – Global Campus, from January 2018 to July 2018.

Medication Management at Acme Medical Center PH4-016

5

provide more of that element without the assistance of a dietician. The new formulation increased the calcium and phosphate concentration to a level outside the acceptable solubility parameters.

3. The general pharmacist who entered the order into the pharmacy system did not recognize the error.

4. The IV solution pharmacist who double-checked the order failed to recognize the error. 5. The TPN solution was prepared off-site by a special-order pharmacy; the preparing

pharmacist failed to recognize the solubility issue. The TPN was prepared as ordered and sent to the medical floor.

6. Because the label verified that the TPN had been formulated as ordered, the bag was hung and TPN started only to have it clog the central line as noted above.

The RCA committee described these events in their report. Dr. Johnson was now pondering what AMC’s next steps should be to ensure that such an event was not repeated.

This document is authorized for use only in Kris Michaelson’s HCM520-WI17D course at Colorado State University – Global Campus, from January 2018 to July 2018.

Medication Management at Acme Medical Center PH4-016

6

Exhibit 1: Overview: Medication Use Process (MUP)

This document is authorized for use only in Kris Michaelson’s HCM520-WI17D course at Colorado State University – Global Campus, from January 2018 to July 2018.

Medication Management at Acme Medical Center PH4-016

7

Exhibit 2: MUP for Pediatric TPN at AMC

Step 1: TPN written on paper order (on weekdays the order is written by registered dietician, and on weekends the order is written by a physician) Step 2: Paper TPN order is delivered from unit to pharmacy Step 3: Pediatric pharmacist reviews TPN order Step 4: Paper order is then taken to IV Pharmacy team. IV room Pharmacist #1 will enter formulation into computer order entry (COE) and check electrolyte balances. Errors: If there are errors noted on the paper order, the Central IV pharmacist must communicate error to prescriber and to pediatric pharmacist

Step 5: IV room pharmacist #2 reviews order now entered into the computer against paper order. Errors: Errors discovered at this point must be communicated to IV Room Pharmacist #1 Step 6: Computer order then goes to outside vendor and TPN is prepared. Vendor pharmacist must check electrolyte composition. Errors: Errors noted by the outside vendor should be communicated to AMC IV room pharmacists. Step 7: TPN prepared at outside vendor Step 8: TPN is delivered back to hospital Step 9: TPN delivered manually to the floor Step 10: Nurse scans barcoded TPN bags Step 11: Administration of medication is documented in electronic MAR

This document is authorized for use only in Kris Michaelson’s HCM520-WI17D course at Colorado State University – Global Campus, from January 2018 to July 2018.

Medication Management at Acme Medical Center PH4-016

8

Exhibit 3: Root Cause Analysis

Level of Analysis Questions Findings Root Cause? Take

Action? What happened?

What are the details of the event? (Brief description)

Did the event pose an immediate danger to patients or staff?

When did the event occur? (Date, day of week, time)

What area/service was impacted? What areas/services/departments

are participating in the root cause analysis?

Why did it happen? What were the most proximate factors? (typically “special cause” variations)

The process or activity in which the event occurred

What are the steps in the process, as designed? (A flow diagram may be helpful here)

Human factors What human factors were relevant to the outcome? (Fatigue of staff, personal problems, unable to focus on task, critical thinking skills needed, rushing to complete task, substance abuse, fatigue or stress)

Organizational factors

What organizational factors are relevant to the outcome?

Equipment factors

How did the equipment performance affect the outcome?

Controllable environmental factors

What factors directly affected the outcome?

This document is authorized for use only in Kris Michaelson’s HCM520-WI17D course at Colorado State University – Global Campus, from January 2018 to July 2018.

Medication Management at Acme Medical Center PH4-016

9

Exhibit 3, continued

Level of Analysis Questions Findings Root Cause? Take

Action? Level of Analysis

Why did that happen? What systems and processes underline those proximate factors? (Common cause variation here may lead to special cause variation in dependent processes.)

Human resource issues

To what degree is staff properly qualified and currently competent for their responsibilities?

How did actual staffing compare with ideal levels?

What are the plans for dealing with contingencies that would tend to reduce effective staffing levels?

To what degree is staff performance in the operant process(es) addressed?

Information management issues

To what degree is all necessary information available when needed? accurate? complete? unambiguous?

To what degree is communication among participants adequate?

Summary of Review Findings/Conclusions (include Root Causes and Contributing Factor/s):

This document is authorized for use only in Kris Michaelson’s HCM520-WI17D course at Colorado State University – Global Campus, from January 2018 to July 2018.