- Email: [email protected]
The Safety Checklist Program: Creating a Culture of Safety in Intensive Care Units
THE JOINT COMMISSION In response to a sentinel event, a safety checklist program was developed to prevent the recurrence of similar errors and help staff proactively examine multiple safety elements in the intensive care unit.
The Safety Checklist Program: Creating a Culture of Safety in Intensive Care Units MARCIA M. PIOTROWSKI, RN, MS DANIEL B. HINSHAW, MD
M
aintaining a safe hospital environment reflects a level of compassion and concern for patient welfare that is as important as any other aspect of competent medical and nursing care. Many institutions have discovered that a single sentinel event can unravel public trust that may take years to rebuild. The Institute of Medicine (IOM) defines patient safety as freedom from accidental injury1 and lists safe care as the first aim for improving the health care system in the 21st century.2 The IOM believes that there are immediate actions that health care organizations can take to promote a safer environment but that the true measure of an institution’s ability to avoid harm will be realized through creating a culture of safety among its employees.3 That culture must allow clinicians to feel secure and even rewarded for seeking and reporting errors and identifying improvement opportunities. The organization must
Marcia M. Piotrowski, RN, MS, is Clinical Risk Manager, Office of the Chief of Staff, VA Ann Arbor Healthcare System (VAAAHS), Ann Arbor, Michigan. Daniel B. Hinshaw, MD, former Chief of Staff at VAAAHS, is Professor of Surgery, University of Michigan Medical School, Ann Arbor, and Staff Physician, Department of Surgery, VAAAHS. Please address requests for reprints to Marcia M. Piotrowski, RN, MS, 2215 Fuller Road, Ann Arbor, MI 48105; phone 734/769-7100, ext 5976; e-mail [email protected]. Copyright © 2002 by the Joint Commission on Accreditation of Healthcare Organizations
306
create a learning environment where errors are valued as opportunities to make systems better and employees can incorporate safety into the fabric of daily practice.3–6 This profound attitudinal shift takes time and deliberate planning. It requires fundamental rethinking of the approach to work, interactions with patients and colleagues, and responsibilities for maintaining a safe environment. Removing a culture of blame does not equate to creating a culture of safety.7 A safe environment is built from the bottom up. A health care institution cannot purport to have a culture of safety without engaging the clinicians who interface with patients. Nurses, physicians, and other support staff must maintain meticulous vigilance over their practice if untoward patient incidents and therapeutic misadventures are to be controlled or eliminated. Clinicians must appreciate that “errors should not be considered as an incurable disease, but rather as preventable phenomena.”8 (p 294) A culture of safety must be supported from the top down. Safety must be a core value of the organization and an explicit corporate priority.9–11 This culture will blossom only with commitment from the chief executive officer (CEO) and other members of the top management team. Executive leaders must “provide strong, clear, and visible attention to safety”1(p 4) and shepherd safety initiatives, providing the resources required for their development.1 Although all patient care settings present safety challenges, observational studies addressing human
JUNE 2002 JOURNAL Copyright 2002 Joint Commission on Accreditation of Healthcare Organizations
JOURNAL
ON
QUALITY IMPROVEMENT
Article-at-a-Glance Background: In 1999 the VA Ann Arbor Healthcare System began a safety checklist program to help build a culture of safety among nurses, respiratory therapists, and unit maintenance providers in the intensive care units (ICUs). Program objectives were to (a) create the opportunity for each participating staff member to view his or her work and unit environment in a broader safety context; (b) establish clear, concise, and measurable standards that staff would identify and value as important safety factors; (c) develop a data collection methodology that would minimize confirmation bias; and (d) correct safety deficits immediately. Data management: Staff measure compliance with safety standards twice daily and record results on a form specifically designed for the project. Data are transferred to a spreadsheet, and graphic presenta-
factors show that that intensive care units (ICUs) are among the highest risk due to the number of activities performed, the complexity of those activities, and the fragile health status of the patients.8,12,13 A recent qualitative study of ICU patients found that their overwhelming need was to feel safe.14 This article describes the efforts undertaken at the VA Ann Arbor Healthcare System (VAAAHS) to develop a performance improvement project that established a culture of safety in medical, surgical, and thoracic ICUs. Program objectives were to ■ create the opportunity for each participating staff member to view her or his work and unit environment in a broader safety context (“What is wrong with the picture?”); ■ establish clear, concise, and measurable standards that staff would identify and value as important safety factors; ■ develop a data collection methodology that would minimize confirmation bias; and ■ correct safety deficits immediately.
Setting The VAAAHS, a Midwestern medical center, is a teaching and comprehensive research facility that has four campuses and serves more than 26,000 patients. It is closely affiliated with a major university and more than 40 other institutions, and it has been providing a comprehensive range of primary, secondary, and tertiary care programs in medicine, surgery, and mental health for nearly 50 years.
VOLUME
28
NUMBER
tions are posted in each ICU. Staff periodically adjust both standards and data collection procedures. Summary: Staff can articulate how the program is making the ICU a safer environment. Nursing response to a recent major error reflects the growth that has occurred since the program’s inception. Safety checks performed by ICU staff are critical in maintaining a constant level of safety. Although the effect on untoward events was not measured, the potential for incidents, including medication and intravenous errors, nosocomial infections, ventilator complications, and restraint complications may be reduced. The program invests bedside clinicians in writing safety standards, creates a partnership between staff and the clinical risk manager, and provides executive leaders an opportunity to demonstrate support of a culture beyond blame.
Sentinel Event Serves as Impetus Before a sentinel event in May 1999 changed clinician complacency, the VAAAHS staff did not visualize the institution as needing major safety improvement initiatives. A nondiabetic ICU patient was inadvertently administered a bolus of regular insulin through his arterial line, resulting in profound brain death. The most likely contributing factor was the improper mingling and storage of multidose vials on top of the medication cart. The loss of life was devastating for the family and the clinicians. In an emotional meeting, the patient’s daughter pleaded with us to make changes that would prevent the recurrence of similar errors. That charge was taken to heart. In response to the untoward event, the chief of staff [D.B.H.] envisioned a safety checklist process (Figure 1, p 308) that would not only resolve difficulties with medication storage but would also proactively examine multiple safety elements in the ICUs. During each shift, nurses would monitor compliance with safety standards. A culture of safety would be integrated into bedside care. Working closely with the clinical risk manager [M.M.P.], the chief of staff met with varied groups of key stakeholders, including unit-level nurse managers, ICU nursing staff, and top management team members. He explained the goals and outlined the process for the program. These exchanges helped to accomplish buyin to the project.
6
Copyright 2002 Joint Commission on Accreditation of Healthcare Organizations
307
THE JOINT COMMISSION Safety Checklist Process Clinical risk manager engages staff and leaders
Staff intermittently update standards and data collection strategies
Staff develop safety standards
Staff measure compliance with safety standards
Staff review areas of weakness and develop plans to improve safety
Clinical risk manager collects data and creates visual displays
Figure 1. This model displays the ongoing process involved in maintaining the safety checklist program.
Developing the Safety Standards Writing the standards was a challenging task. They needed to be terse, readily understood, and consistently interpreted. The content had to reflect safety elements that the nurses collectively believed were critical. Further, there was an ethical obligation to initiate the program quickly. Finally, it was recognized that this project, although moderately complex and requiring a high degree of oversight and support, was not a research study but rather a quality management tool for enhancing safety. The goal was to foster recognition among staff that they are pivotal in maintaining standards that potentially prevent major disabilities and save lives. It was decided to forego a period of baseline data collection. Rather than wait to
308
have a polished product, VAAAHS rapidly developed a draft checklist, and conducted a trial of that form with two nurses over a brief period of 2 days. Feedback was used to revise the form and then trial again. When the tool was sufficiently refined, all ICU nurses used it for a longer period of time. To write the original safety standards, the clinical risk manager stationed herself in an ICU break room for several hours, permitting an active exchange of ideas. Nurses joined the conversation as patient care duties allowed. Two days later the nurses tried the first draft of the safety checklist. Information on whether each standard was met or unmet was recorded on a form designed specifically for this project. This paper tool included the name of the unit, the date, individ-
JUNE 2002 JOURNAL Copyright 2002 Joint Commission on Accreditation of Healthcare Organizations
JOURNAL
ON
QUALITY IMPROVEMENT
ual standards, and boxes where the nurse could quickly check either yes or no to indicate compliance with each safety element. During the next five weeks, six versions of the form were created and used. The number of standards and the intent of each standard remained constant, but the wording was increasingly refined. The nurses felt that they owned the content. All changes reflected feedback from ICU staff. The nurses encouraged the clinical risk manager to involve respiratory therapists and unit maintenance personnel. Members of these support services performed functions that were critical to patient safety. Both the leaders and staff of these services were anxious to be included. A similar methodology was used to develop two additional lists. Safety standards included in all three checklists are summarized in Table 1 (p 310). This comprehensive list of standards reflects safety elements that are either currently used or have been measured at some point since the initiation of the program. Selected standards have been tracked throughout the entire length of the program; not every standard is measured daily. The evolution of the measurement process is described in the following section.
Modifying the Safety Checklist Program The safety checklist program began in Autumn 1999. Today the process of revising checklists continues, and it is anticipated that it will remain an essential component of a viable program. Simply tracking a safety standard increases the likelihood of compliance. There have been three configurations of nursing safety standards (Table 2, p 311) as the program has evolved. Changes in the safety elements and methods of data collection reflect increased sophistication and growth in the program. Initially, nurses examined 26 safety items that were divided onto three separate daily data collection forms. Each form included a place to record compliance by standard for each patient room. Three nurses were each accountable for one of the forms. Although dividing the responsibility among staff reduced the time burden, it still took individual nurses about 10 minutes to accurately evaluate compliance. On busy days it was difficult to find even a short block of time, although the nurses were committed and did make substantial efforts to complete the checklists. By the
VOLUME
28
NUMBER
end of the first year (Autumn 2000), most of the standards were being consistently met, with an average compliance rate of greater than 90%. The nurses felt that it was time to revise the lists, so they selected fewer elements to reduce the time burden yet maintain the goal of enhancing a culture of safety through heightened awareness of critical safety factors. Four items were chosen for daily measurement on all ICU patients. These core elements were ■ intravenous (IV) pumps labeled with infusion name and matched solution in IV bag; ■ bioclusive dressings dated and timed on arterial, central, and peripheral IV insertion sites, and each dressing changed within the past 72 hours; ■ gauze dressings (gauze next to puncture site even if a bioclusive dressing was also applied) dated and timed on arterial, central, and peripheral IV insertion sites, and each dressing changed within the past 24 hours; and ■ an identification (ID) band on each patient’s wrist. To appreciate the selection of these particular safety components one must understand the historical framework of the medical center’s ICU environment. The occurrence of IV errors and IV line infections was particularly worrisome to staff. To overcome the former concern, nurses had implemented a redundant system of labeling with the IV solution displayed at three sites: on the bag, on the tubing near the point of skin insertion, and on the pump. Nurses wanted to make certain this effective procedure did not extinguish over time. IV dressing changes had also been an area of ongoing concern that the staff believed required daily vigilance. Finally, ID bands were critical for identification, particularly when patients were intubated or experiencing a reduced level of consciousness. Intermittently, the name band was cut off the wrist and taped to the bed rather than replaced. This lax practice needed to cease. In addition to the core elements, five other standards addressing compliance with isolation protocols and restraint use were labeled as “intermittent,” to be checked daily if applicable to the patient. The remaining items were divided between four separate lists. A different list was checked during one month (January, April, July, and October) each quarter. During the course of the second year of the program, a team was formed to improve ventilator care. As an outgrowth of that effort, seven additional safety elements affecting
6
Copyright 2002 Joint Commission on Accreditation of Healthcare Organizations
309
THE JOINT COMMISSION Table 1. Safety Standards Measured by Each Discipline* Nurses
Medications ■ Medication carts locked ■ Medication vials secured inside medication carts ■ Patient rooms free of medication vials, needles, and syringes ■ Double signature on medication record for required medications ■ Complete documentation of prn medications Ventilators Sedation scale documented every 4 hours
■
Intravenous Solutions ■ IV pump labeled with infusion name ■ IV solutions and rates correct ■ IV sites pain free and without erythema ■ IV tubing dated, timed when hung, and changed within 72 hours Equipment Bedside alarms on for continuous monitoring of EKG and arterial blood pressure ■ Two EKG leads monitored, with one being V lead ■ NG canister dated, timed, and changed within 72 hours ■ Electrodes changed every 48 hours ■
Environment Sterile/clean drawers, free of patient personal belongings and contaminated supplies ■ Adherence to isolation protocols ■
Restraint Restraint device(s) properly applied ■ Restraint safety flow sheet completed ■ Physician restraint order completed ■ Physician face-to-face restraint assessment completed ■
Other Mouth care every 4 hours ■ Pain scale documented every 4 hours ■ All dressings dated and changed according to policy ■ Patient turned every 2 hours ■ Patient up in chair at least once daily ■ Head of bed up 30 degrees ■ ID band on patient ■ Open solution bottles dated and timed; no outdated solutions ■
Respiratory Therapists
Medications ■ Respiratory therapy cart free of medications, both inside and on top ■ Patient rooms free of respiratory therapy medications (Exception: patient in isolation) ■ Medication administration record signed Ventilators Ventilator circuit changed within 7 days; documented on ventilator flow sheet ■ Ventilator volume and pressure alarms on ■ Cuff pressures documented on ventilator flow sheet ■ Patient assessment documented on back of ventilator flow sheet ■ Endotracheal tube position documented each shift ■
■
Endotracheal tube tape changed every 48 hours and prn; documented on ventilator flow sheet
Oxygen Oxygen equipment (eg, mask, cannula) changed weekly and documented on oxygen flow sheet ■ Oxygen cylinders stored in appropriate stands ■
Other Ambu bag in room (for both ventilator and nonventilator patients) and attached to O2 ■ Tracheostomy care done and ties changed; documented on ventilator flow sheet ■ Pulmonary/oral secretions suction tubing and canister dated, timed, and changed within 48 hours ■
Unit Maintenance Providers
Environment ■ Clear passage in hallway ■ Floors free of spills ■ Electrical cords secured and not obstructing environment ■ Sharps containers no more than 2/3 full; no item protruding from container ■ Linen room door closed and latched ■ Dirty utility room door closed and latched ■ Supply rooms free of corrugated cardboard ■ Housekeeping cart clean from previous shift
■ ■ ■ ■
Housekeeping closet free of clutter Chemicals properly stored (securely locked and not stored above head level) Unit free of employee food and drink Trash container lids clear of items
Equipment ■ Nonfunctioning equipment labeled; copy of work order attached ■ Wet vacuum available in closet ■ Brakes on wheelchairs properly functioning ■ Refrigerator temperature monitored
* prn, as needed; IV, intravenous; NG, nasogastric; EKG, electrocardiogram; ID, identification.
310
JUNE 2002 JOURNAL Copyright 2002 Joint Commission on Accreditation of Healthcare Organizations
JOURNAL
ON
QUALITY IMPROVEMENT
Table 2. The Three Configurations of the Nursing Safety Standards* ■ ■
■
26 daily standards, divided into a three-part list (Oct 1999–Sep 2000) 4 daily standards + 5 intermittent standards + 5 standards rotating each quarter + 7 ventilator standards (36 total standards; Oct 2000– Jul 2001) 2 daily standards (rotate weekly over 10-week cycle) + 5 intermittent standards (25 total standards; Aug 2001–present)
* Table 1 (p 310) is a comprehensive list of safety standards used at some point during the safety checklist program. Not all of the standards are currently employed.
ventilator complications were added. The ventilator standards included items such as changing endotracheal tube position, providing frequent mouth care, and monitoring sedation levels. The process for monitoring safety standards was further refined in August 2001. Two new standards are posted weekly. Instead of daily forms, a weekly data collection tool is now placed in each room. The five intermittent standards are included on the weekly form. There are a total of 10 unique sheets. Twice daily, at change of shift, the offgoing and oncoming nurses jointly document compliance. The technique is accomplishing several positive practice changes: ■ It removes the opportunity for confirmation bias, defined as the predisposition to “accept information that agrees with our hypothesis and reject information that does not.”15 In the context of safety, this translates to seeing what one expects to see. For example, while replacing tubing on a bank of IVs, a nurse may accidentally reconnect two lines improperly. The nurse may not recognize the error because he or she expects to see the IVs reassembled correctly. However, a colleague may readily unearth the mistake. When the program began, the nurses were asked to check rooms other than their own to avoid the risk of confirmation error. Having two nurses perform the task in tandem greatly reduces the possibility of following the wrong procedure. ■ It serves as an additional impetus for bedside rounds. ■ It essentially eliminates any time burden associated with earlier versions of the checklist. The nurses incorporate the safety checks into their change of shift report, recording compliance information at that time.
VOLUME
28
NUMBER
Weekly rotation of standards keeps the program fresh while allowing each standard to be measured during several weeks interspersed throughout the year. ■ Reducing the number of items being monitored increases compliance with twice-daily checks. ■
Data Management Quick turnaround of information is a key to the program’s success. The program not only provides timely feedback to staff but also indicates that the performance improvement service is a responsive partner in this venture. To accomplish timely posting of results, students from a nearby university were hired on a part-time basis to set up spreadsheets, transfer data from the forms completed on the units to the spreadsheets, and create three visual displays that are exhibited in each ICU. These displays are regularly updated. Figure 2 (p 312) presents a sample bar graph that demonstrates the percentage of compliance with each standard. Respiratory therapists and unit maintenance providers each maintain an established set of 14 safety standards. Although the two groups look at different safety elements specific to the responsibilities of the service, the standards do not change from day to day. The data on percentage of compliance with individual standards for respiratory therapists and unit maintenance providers is summarized and posted monthly. As described earlier, nurses look at 2 new standards each week, with 20 standards measured during a 10-week period. The results are posted at the conclusion of each 10-week cycle. Figure 3 (p 312) presents a second sample bar graph that depicts the number of times each standard is unmet. Because compliance is generally high, aggregate data on the number of incidents of noncompliance for respiratory therapists and unit maintenance providers are posted quarterly rather than monthly. For nurses these data are updated and displayed for each 10-week cycle of standard measurement. Finally, Table 3 (p 313) provides a representative sample of individual incidents of noncompliance with standards. Tables and graphs are posted in the ICU for nurses, respiratory therapists, and unit maintenance providers. Reviewed in combination, the displays permit staff to gather both general and detailed information on the results of their checklist measurements.
6
Copyright 2002 Joint Commission on Accreditation of Healthcare Organizations
311
THE JOINT COMMISSION Representative Sample of Compliance with Individual Safety Standards, Jan–Jun 2001 Registered Nurses’ Compliance with Safety Standards for Ventilated Patients in the Thoracic Intensive Care Unit 100%
100%
100%
100%
% Compliance with Standards
90% 89%
93%
77%
70% 60%
89%
88%
80%
67%
64%
Jan ’01 Feb ’01 Mar ’01 Apr ’01 May ’01 Jun ’01
58%
50% 40% 30%
29%
20% 10% 0% Ramsey Scale Documented Mouth Care Done Ventilated Patient Standards
Figure 2. This representative sample of a bar graph depicts the percentage of compliance with individual standards. Graphs like this one are posted on each unit.
Representative Sample of a Number of Incidents of Noncompliance Noncompliance with Standards for Respiratory Therapists in the Intensive Care Unit
Oxygen equipment changed within 72 hours
12
Ambu bag in room
4
Pulmonary/oral secretion tubing and canister changed within 48 hours
7
5
0
7
1st quarter 2001 2nd quarter 2001 3rd quarter 2001
10
10
5
4
10
4
15
20
25
Number of Incidents of Noncompliance Figure 3. This representative sample of a bar graph depicts the number of incidents of noncompliance with individual safety standards. Graphs
like this one are posted on each unit.
312
JUNE 2002 JOURNAL Copyright 2002 Joint Commission on Accreditation of Healthcare Organizations
JOURNAL
ON
QUALITY IMPROVEMENT
Table 3. Safety Tracking for Registered Nurses in the Thoracic Intensive Care Unit* Date
Room
Topic
Shift
Issue
10/13/01
6A
M
nights
Double signature missing on medication administration record (indicates double-check of high-risk IV infusions)
10/13/01
6B
O
days
Mouth care not documented every 4 hours for ventilated patient
10/16/01
6F
O
nights
Mouth care not documented every 4 hours for ventilated patient
10/18/01
6A
O
days
Pain scale not documented every 4 hours
10/21/01
6E
IV
days
IV pumps and/or tubing not labeled
10/22/01
6C
O
days
Pain scale not documented every 4 hours
10/26/01
6B
O
nights
Pain scale not documented every 4 hours
10/27/01
6F
V
nights
Sedation scale not recorded every 4 hours for ventilated patient
10/27/01
6C
V
nights
Sedation scale not recorded every 4 hours for ventilated patient
10/30/01
6D
O
days
Pain scale not documented every 4 hours
11/1/01
6B
O
nights
Pain scale not documented every 4 hours
11/1/01
6A
V
days
IV pumps and/or tubing not labeled
11/2/01
6D
M
days
Documentation of prn medications not completed
11/5/01
6D
M
nights
Documentation of prn medications not completed
* Dates and room numbers have been changed to protect confidentiality. M, medication; IV, intravenous; O, other; V, ventilator; prn, as needed.
Outcomes Since the inception of the checklist program, thousands of measurements have been taken that reflect either compliance or noncompliance with the safety standards. It is challenging to quantify the safety advances because the checklist configuration has been altered several times, compliance within each discipline (for example, nurses, respiratory therapists, unit maintenance personnel) varies by individual ICU, and even within each ICU compliance with several standards has varied over time. Moreover, a few of the standards were originally measured by one discipline but are currently measured by another group. For example, nurses initially collected data on the dating and replacement of pulmonary suction tubing and canisters. Now respiratory therapists gather that information. Improved compliance with these select standards, however, does not address the truly critical outcomes. These outcomes may be best pursued by seeking the answers to five questions: 1. Have there been recurrences of the same type of medication error, involving nonsecure medications on carts, which led to the original sentinel event? 2. Have there been similar medication errors?
VOLUME
28
NUMBER
3. Have patient deaths or cases of permanent morbidity linked to safety deficits occurred since the inception of the program? 4. Have staff been able to transfer their learning and experience about a specific problem to the general concept of creating a safe environment instead of solely focusing on discrete safety elements? 5. Do staff attitudes regarding the occurrence of error differ as compared with before the program’s inception? There have been no recurrences of error linked to improper storage of medications on top of carts. However, in the past 2 years there have been two errors with a significant risk of morbidity or mortality. Neither incident resulted in permanent injury. The first error occurred during the early months of program implementation. IV potassium was infused through an arterial line for approximately 8 hours. One nurse improperly connected the line just before midnight, but the nurse caring for the patient during the following shift failed to perform the prescribed safety check. A third nurse who entered duty at 8:00 AM discovered the improper connection, which alerted the staff to the importance of the checklist. The second incident took place recently and involved hanging via IV piggyback (IVPB) a neuromuscular blocker on
6
Copyright 2002 Joint Commission on Accreditation of Healthcare Organizations
313
THE JOINT COMMISSION a patient being weaned from the ventilator. The IVPB was intended for another patient and was properly labeled. A major proximal cause of the error was unsafe separation and storage of IVPB medications before administration. Although in retrospect the risky practice was transparent, before this event the staff had not associated the safety concerns connected with multidose vial handling with other forms of medication administration. The transfer of the learning experience occurred after the event and is an area where there is room for continued growth. The fourth question deals with staff attitudes toward safety and the incorporation of the “big safety picture” into daily practice. Indirect measures indicate that the program is achieving that goal. Staff are invested in the initiative, as evidenced by their ongoing revisions of both the safety standards and the methods of data collection. The clinical risk manager regularly queries staff about the value of the program. Nurses, respiratory therapists, and unit maintenance providers are able to define a culture of safety and describe the positive impacts that the program has had on unit-level safety. Reaction to the recent IVPB error reflects a change in attitude since the time of the initial sentinel event. The nurse involved immediately disclosed the error to her supervisor and was well supported. Systems weaknesses were reviewed that day, and several changes were made, to reduce the likelihood of recurrence. Compliance with those changes is being monitored through the checklist program and will remain a daily safety standard until the practices are deeply ingrained in the routines of pharmacists and nurses. This response was a major positive shift from the initial reaction to the original sentinel event that engendered the safety checklist. It speaks eloquently of the growth that has transpired in this staff. Occasionally there is slippage in compliance with select safety standards. Posting the data and bringing weaknesses to the attention of the staff has corrected most difficulties. Supervisory support is critical. Unit leaders must point out the problems without blaming individuals or pitting one shift against another. The leaders must also guide staff in formulating solutions.
Lessons Learned Initially, many staff members believed that this project was a
Staff trust is critical to program success.
314
means of challenging and/or measuring their competence. They saw the vocalized intent of using the checklist to improve patient safety as a thinly veiled disguise for assessing individual performance. This deeply rooted conviction was often not openly expressed but nonetheless limited early participation. It persisted for a time, despite efforts at education, including viewing the video Beyond Blame16 and group discussions of a systems approach to medical error. The facility had a cultural history that was often quick to blame individuals deemed responsible for error. It was naïve to think that this culture could change within weeks. Staff observation of leadership behaviors was critical. Over the course of several months, clinicians began to trust leadership actions. After a shaky start-up, nurses, respiratory therapists, and unit maintenance personnel began to understand that less-than-perfect performance did not reflect negatively on professional practice. Partnerships create vital programs. The safety checklist program is a team effort between staff, unitlevel leadership, the chief of staff, and the clinical risk manager. These partners recognize their interdependence and celebrate their accomplishments. Team members acknowledge that they need to remain faithful to their portion of the agreement. Nurses and ancillary staff must complete the safety checks, and the clinical risk manger is responsible for technical support and rapid turnaround of aggregate data. Maintaining the program takes effort. The initial excitement, concerns, and adjustments to the checklist have faded, but there is a need to keep the program going over the years. Unless carefully nurtured, the momentum will diminish. To provide ongoing support, the clinical risk manager regularly visits the ICUs and listens to staff, both one-on-one and in small groups. During intermittent group meetings, checklist data are shared and the group discusses how to enhance the effort. Finally, clinicians remain excited if leadership remains committed and enthusiastic.
Future Directions Now that the ICU safety checklist program is well established, discussions have been held on how to enhance this initiative. The program will be widened to include input from patients and families.17–18 Stakeholders will be informed about the initiative through
JUNE 2002 JOURNAL Copyright 2002 Joint Commission on Accreditation of Healthcare Organizations
JOURNAL
ON
QUALITY IMPROVEMENT
a brochure. A focus group composed of several patients and family members, staff clinicians, and the clinical risk manager will be formed to elicit ideas for consumer involvement. The second thrust is to expand the program throughout the organization. Giving the ICU staff an opportunity to serve as expert consultants to other areas within the organization that are attempting to develop their own safety programs will give the pioneers much-deserved recognition and also ensure their continued enthusiasm for their own program. A tentative schedule for spread has been established, beginning with mental health and the operating room. Lessons learned through these initial years should smooth introduction of the process, thus reducing start-up time.
Summary Twice-daily safety checks performed by staff in direct care roles are critical in maintaining a constant level of safety. Although the safety standards need to be sound and reflect aspects of care that affect safety, achieving
the “perfect” set of standards is not the core goal. The critical objective is to have clinicians involved in hands-on care incorporate a culture of safety into their daily practice; it is to stand back from the discrete details of their complex work and contemplate the broader picture of how they can improve safety. The project has several key advantages: ■ Bedside clinicians are responsible for and invested in developing safety standards and measuring their practice against those standards; ■ A solid partnership is forged between clinicians and the clinical risk manager, a vital component for effective safety efforts; and ■ Top organizational leaders are provided an opportunity to demonstrate their promise of creating a culture beyond blame. Although not measured in this program, the potential for untoward events, including medication and IV errors, nosocomial infections, ventilator complications, and restraint complications, may be reduced through maintaining a unit-level safety checklist. J
References 1. Institute of Medicine: To Err Is Human: Building a Safer Health System. Washington, DC: National Academy Press, 2000. 2. Institute of Medicine: Crossing the Quality Chasm: A New Health System for the 21st Century. Washington, DC: National Academy Press, 2001. 3. Loeb J: To err is human: An interview with the Institute of Medicine’s Linda Kohn. Jt Comm J Qual Improv 26:227–234, 2000. 4. O’Leary D: Organizational evaluation and a culture of safety. In Proceedings of Enhancing Patient Safety and Reducing Errors in Health Care, Nov 8–10, 1998. Chicago: National Patient Safety Foundation, American Medical Association, 1999, pp 34–37. 5. Berman S: The AMA clinical quality improvement forum on addressing patient safety. Jt Comm J Qual Improv 26:428–433, 2000.
VOLUME
28
NUMBER
6. Phillips DF: “New Look” reflects changing style of patient safety enhancement. JAMA 281: 217–219, 1999. 7. Leape LL (moderator): Creating a culture of safety. In Proceedings of Enhancing Patient Safety and Reducing Errors in Health Care, Nov 8–10, 1998. Chicago: National Patient Safety Foundation, American Medical Association, 1999, pp 23–24. 8. Donchin Y, et al: A look into the nature and causes of human errors in the intensive care unit. Crit Care Med 23:294–300, 1995. 9. Reinertsen JL: Network management and a culture of safety. In Proceedings of Enhancing Patient Safety and Reducing Errors in Health Care, Nov 8–10, 1998. Chicago: National Patient Safety Foundation, American Medical Association, 1999, pp 28–30.
10. Kizer K: Large system change and a culture of safety. In Proceedings of Enhancing Patient Safety and Reducing Errors in Health Care, Nov 8–10, 1998. Chicago: National Patient Safety Foundation, American Medical Association, 1999, pp 31–33. 11. Leape LL: Can we make health care safe? In Reducing Medication Errors and Improving Patient Safety: Success Stories from the Front Lines of Medicine (report). Accelerating Change Today (conference). National Coalition on Health Care (Washington, DC) and the Institute for Healthcare Improvement (Boston), Feb 2000, pp 2–3. 12. Giraud T, et al: Iatrogenic complications in adult intensive care units: A prospective two-center study. Crit Care Med 21:40–51, 1993. 13. Rubins HB, Moskowitz MA: Complications of care in a medical
intensive care unit. J Gen Intern Med 5:104–109, 1990. 14. Hupcey JE: Feeling safe: The psychosocial needs of ICU patients. J Nurs Sch 32:361–367, 2000. 15. Cohen M. Reducing medication errors: We already know what to do! (presentation at Veterans Health Administration and Institute for Healthcare Improvement Collaborative on Reducing Adverse Drug Events). Baltimore, Apr 11, 2000. 16. Bridge Medical: Beyond Blame (videotape) documentary. Solana Beach, CA, 1997. 17. Turning medical error tragedy into opportunity. Executive Solutions for Healthcare Management 3:9–11, Jan 2000. 18. Want a savvy participant in your error-prevention program? Put a consumer on your team! ISMP Medication Safety Alert 5(10):1, 2000.
6
Copyright 2002 Joint Commission on Accreditation of Healthcare Organizations
315
The Safety Checklist Program: Creating a Culture of Safety in Intensive Care Units MARCIA M. PIOTROWSKI, RN, MS DANIEL B. HINSHAW, MD
M
aintaining a safe hospital environment reflects a level of compassion and concern for patient welfare that is as important as any other aspect of competent medical and nursing care. Many institutions have discovered that a single sentinel event can unravel public trust that may take years to rebuild. The Institute of Medicine (IOM) defines patient safety as freedom from accidental injury1 and lists safe care as the first aim for improving the health care system in the 21st century.2 The IOM believes that there are immediate actions that health care organizations can take to promote a safer environment but that the true measure of an institution’s ability to avoid harm will be realized through creating a culture of safety among its employees.3 That culture must allow clinicians to feel secure and even rewarded for seeking and reporting errors and identifying improvement opportunities. The organization must
Marcia M. Piotrowski, RN, MS, is Clinical Risk Manager, Office of the Chief of Staff, VA Ann Arbor Healthcare System (VAAAHS), Ann Arbor, Michigan. Daniel B. Hinshaw, MD, former Chief of Staff at VAAAHS, is Professor of Surgery, University of Michigan Medical School, Ann Arbor, and Staff Physician, Department of Surgery, VAAAHS. Please address requests for reprints to Marcia M. Piotrowski, RN, MS, 2215 Fuller Road, Ann Arbor, MI 48105; phone 734/769-7100, ext 5976; e-mail [email protected]. Copyright © 2002 by the Joint Commission on Accreditation of Healthcare Organizations
306
create a learning environment where errors are valued as opportunities to make systems better and employees can incorporate safety into the fabric of daily practice.3–6 This profound attitudinal shift takes time and deliberate planning. It requires fundamental rethinking of the approach to work, interactions with patients and colleagues, and responsibilities for maintaining a safe environment. Removing a culture of blame does not equate to creating a culture of safety.7 A safe environment is built from the bottom up. A health care institution cannot purport to have a culture of safety without engaging the clinicians who interface with patients. Nurses, physicians, and other support staff must maintain meticulous vigilance over their practice if untoward patient incidents and therapeutic misadventures are to be controlled or eliminated. Clinicians must appreciate that “errors should not be considered as an incurable disease, but rather as preventable phenomena.”8 (p 294) A culture of safety must be supported from the top down. Safety must be a core value of the organization and an explicit corporate priority.9–11 This culture will blossom only with commitment from the chief executive officer (CEO) and other members of the top management team. Executive leaders must “provide strong, clear, and visible attention to safety”1(p 4) and shepherd safety initiatives, providing the resources required for their development.1 Although all patient care settings present safety challenges, observational studies addressing human
JUNE 2002 JOURNAL Copyright 2002 Joint Commission on Accreditation of Healthcare Organizations
JOURNAL
ON
QUALITY IMPROVEMENT
Article-at-a-Glance Background: In 1999 the VA Ann Arbor Healthcare System began a safety checklist program to help build a culture of safety among nurses, respiratory therapists, and unit maintenance providers in the intensive care units (ICUs). Program objectives were to (a) create the opportunity for each participating staff member to view his or her work and unit environment in a broader safety context; (b) establish clear, concise, and measurable standards that staff would identify and value as important safety factors; (c) develop a data collection methodology that would minimize confirmation bias; and (d) correct safety deficits immediately. Data management: Staff measure compliance with safety standards twice daily and record results on a form specifically designed for the project. Data are transferred to a spreadsheet, and graphic presenta-
factors show that that intensive care units (ICUs) are among the highest risk due to the number of activities performed, the complexity of those activities, and the fragile health status of the patients.8,12,13 A recent qualitative study of ICU patients found that their overwhelming need was to feel safe.14 This article describes the efforts undertaken at the VA Ann Arbor Healthcare System (VAAAHS) to develop a performance improvement project that established a culture of safety in medical, surgical, and thoracic ICUs. Program objectives were to ■ create the opportunity for each participating staff member to view her or his work and unit environment in a broader safety context (“What is wrong with the picture?”); ■ establish clear, concise, and measurable standards that staff would identify and value as important safety factors; ■ develop a data collection methodology that would minimize confirmation bias; and ■ correct safety deficits immediately.
Setting The VAAAHS, a Midwestern medical center, is a teaching and comprehensive research facility that has four campuses and serves more than 26,000 patients. It is closely affiliated with a major university and more than 40 other institutions, and it has been providing a comprehensive range of primary, secondary, and tertiary care programs in medicine, surgery, and mental health for nearly 50 years.
VOLUME
28
NUMBER
tions are posted in each ICU. Staff periodically adjust both standards and data collection procedures. Summary: Staff can articulate how the program is making the ICU a safer environment. Nursing response to a recent major error reflects the growth that has occurred since the program’s inception. Safety checks performed by ICU staff are critical in maintaining a constant level of safety. Although the effect on untoward events was not measured, the potential for incidents, including medication and intravenous errors, nosocomial infections, ventilator complications, and restraint complications may be reduced. The program invests bedside clinicians in writing safety standards, creates a partnership between staff and the clinical risk manager, and provides executive leaders an opportunity to demonstrate support of a culture beyond blame.
Sentinel Event Serves as Impetus Before a sentinel event in May 1999 changed clinician complacency, the VAAAHS staff did not visualize the institution as needing major safety improvement initiatives. A nondiabetic ICU patient was inadvertently administered a bolus of regular insulin through his arterial line, resulting in profound brain death. The most likely contributing factor was the improper mingling and storage of multidose vials on top of the medication cart. The loss of life was devastating for the family and the clinicians. In an emotional meeting, the patient’s daughter pleaded with us to make changes that would prevent the recurrence of similar errors. That charge was taken to heart. In response to the untoward event, the chief of staff [D.B.H.] envisioned a safety checklist process (Figure 1, p 308) that would not only resolve difficulties with medication storage but would also proactively examine multiple safety elements in the ICUs. During each shift, nurses would monitor compliance with safety standards. A culture of safety would be integrated into bedside care. Working closely with the clinical risk manager [M.M.P.], the chief of staff met with varied groups of key stakeholders, including unit-level nurse managers, ICU nursing staff, and top management team members. He explained the goals and outlined the process for the program. These exchanges helped to accomplish buyin to the project.
6
Copyright 2002 Joint Commission on Accreditation of Healthcare Organizations
307
THE JOINT COMMISSION Safety Checklist Process Clinical risk manager engages staff and leaders
Staff intermittently update standards and data collection strategies
Staff develop safety standards
Staff measure compliance with safety standards
Staff review areas of weakness and develop plans to improve safety
Clinical risk manager collects data and creates visual displays
Figure 1. This model displays the ongoing process involved in maintaining the safety checklist program.
Developing the Safety Standards Writing the standards was a challenging task. They needed to be terse, readily understood, and consistently interpreted. The content had to reflect safety elements that the nurses collectively believed were critical. Further, there was an ethical obligation to initiate the program quickly. Finally, it was recognized that this project, although moderately complex and requiring a high degree of oversight and support, was not a research study but rather a quality management tool for enhancing safety. The goal was to foster recognition among staff that they are pivotal in maintaining standards that potentially prevent major disabilities and save lives. It was decided to forego a period of baseline data collection. Rather than wait to
308
have a polished product, VAAAHS rapidly developed a draft checklist, and conducted a trial of that form with two nurses over a brief period of 2 days. Feedback was used to revise the form and then trial again. When the tool was sufficiently refined, all ICU nurses used it for a longer period of time. To write the original safety standards, the clinical risk manager stationed herself in an ICU break room for several hours, permitting an active exchange of ideas. Nurses joined the conversation as patient care duties allowed. Two days later the nurses tried the first draft of the safety checklist. Information on whether each standard was met or unmet was recorded on a form designed specifically for this project. This paper tool included the name of the unit, the date, individ-
JUNE 2002 JOURNAL Copyright 2002 Joint Commission on Accreditation of Healthcare Organizations
JOURNAL
ON
QUALITY IMPROVEMENT
ual standards, and boxes where the nurse could quickly check either yes or no to indicate compliance with each safety element. During the next five weeks, six versions of the form were created and used. The number of standards and the intent of each standard remained constant, but the wording was increasingly refined. The nurses felt that they owned the content. All changes reflected feedback from ICU staff. The nurses encouraged the clinical risk manager to involve respiratory therapists and unit maintenance personnel. Members of these support services performed functions that were critical to patient safety. Both the leaders and staff of these services were anxious to be included. A similar methodology was used to develop two additional lists. Safety standards included in all three checklists are summarized in Table 1 (p 310). This comprehensive list of standards reflects safety elements that are either currently used or have been measured at some point since the initiation of the program. Selected standards have been tracked throughout the entire length of the program; not every standard is measured daily. The evolution of the measurement process is described in the following section.
Modifying the Safety Checklist Program The safety checklist program began in Autumn 1999. Today the process of revising checklists continues, and it is anticipated that it will remain an essential component of a viable program. Simply tracking a safety standard increases the likelihood of compliance. There have been three configurations of nursing safety standards (Table 2, p 311) as the program has evolved. Changes in the safety elements and methods of data collection reflect increased sophistication and growth in the program. Initially, nurses examined 26 safety items that were divided onto three separate daily data collection forms. Each form included a place to record compliance by standard for each patient room. Three nurses were each accountable for one of the forms. Although dividing the responsibility among staff reduced the time burden, it still took individual nurses about 10 minutes to accurately evaluate compliance. On busy days it was difficult to find even a short block of time, although the nurses were committed and did make substantial efforts to complete the checklists. By the
VOLUME
28
NUMBER
end of the first year (Autumn 2000), most of the standards were being consistently met, with an average compliance rate of greater than 90%. The nurses felt that it was time to revise the lists, so they selected fewer elements to reduce the time burden yet maintain the goal of enhancing a culture of safety through heightened awareness of critical safety factors. Four items were chosen for daily measurement on all ICU patients. These core elements were ■ intravenous (IV) pumps labeled with infusion name and matched solution in IV bag; ■ bioclusive dressings dated and timed on arterial, central, and peripheral IV insertion sites, and each dressing changed within the past 72 hours; ■ gauze dressings (gauze next to puncture site even if a bioclusive dressing was also applied) dated and timed on arterial, central, and peripheral IV insertion sites, and each dressing changed within the past 24 hours; and ■ an identification (ID) band on each patient’s wrist. To appreciate the selection of these particular safety components one must understand the historical framework of the medical center’s ICU environment. The occurrence of IV errors and IV line infections was particularly worrisome to staff. To overcome the former concern, nurses had implemented a redundant system of labeling with the IV solution displayed at three sites: on the bag, on the tubing near the point of skin insertion, and on the pump. Nurses wanted to make certain this effective procedure did not extinguish over time. IV dressing changes had also been an area of ongoing concern that the staff believed required daily vigilance. Finally, ID bands were critical for identification, particularly when patients were intubated or experiencing a reduced level of consciousness. Intermittently, the name band was cut off the wrist and taped to the bed rather than replaced. This lax practice needed to cease. In addition to the core elements, five other standards addressing compliance with isolation protocols and restraint use were labeled as “intermittent,” to be checked daily if applicable to the patient. The remaining items were divided between four separate lists. A different list was checked during one month (January, April, July, and October) each quarter. During the course of the second year of the program, a team was formed to improve ventilator care. As an outgrowth of that effort, seven additional safety elements affecting
6
Copyright 2002 Joint Commission on Accreditation of Healthcare Organizations
309
THE JOINT COMMISSION Table 1. Safety Standards Measured by Each Discipline* Nurses
Medications ■ Medication carts locked ■ Medication vials secured inside medication carts ■ Patient rooms free of medication vials, needles, and syringes ■ Double signature on medication record for required medications ■ Complete documentation of prn medications Ventilators Sedation scale documented every 4 hours
■
Intravenous Solutions ■ IV pump labeled with infusion name ■ IV solutions and rates correct ■ IV sites pain free and without erythema ■ IV tubing dated, timed when hung, and changed within 72 hours Equipment Bedside alarms on for continuous monitoring of EKG and arterial blood pressure ■ Two EKG leads monitored, with one being V lead ■ NG canister dated, timed, and changed within 72 hours ■ Electrodes changed every 48 hours ■
Environment Sterile/clean drawers, free of patient personal belongings and contaminated supplies ■ Adherence to isolation protocols ■
Restraint Restraint device(s) properly applied ■ Restraint safety flow sheet completed ■ Physician restraint order completed ■ Physician face-to-face restraint assessment completed ■
Other Mouth care every 4 hours ■ Pain scale documented every 4 hours ■ All dressings dated and changed according to policy ■ Patient turned every 2 hours ■ Patient up in chair at least once daily ■ Head of bed up 30 degrees ■ ID band on patient ■ Open solution bottles dated and timed; no outdated solutions ■
Respiratory Therapists
Medications ■ Respiratory therapy cart free of medications, both inside and on top ■ Patient rooms free of respiratory therapy medications (Exception: patient in isolation) ■ Medication administration record signed Ventilators Ventilator circuit changed within 7 days; documented on ventilator flow sheet ■ Ventilator volume and pressure alarms on ■ Cuff pressures documented on ventilator flow sheet ■ Patient assessment documented on back of ventilator flow sheet ■ Endotracheal tube position documented each shift ■
■
Endotracheal tube tape changed every 48 hours and prn; documented on ventilator flow sheet
Oxygen Oxygen equipment (eg, mask, cannula) changed weekly and documented on oxygen flow sheet ■ Oxygen cylinders stored in appropriate stands ■
Other Ambu bag in room (for both ventilator and nonventilator patients) and attached to O2 ■ Tracheostomy care done and ties changed; documented on ventilator flow sheet ■ Pulmonary/oral secretions suction tubing and canister dated, timed, and changed within 48 hours ■
Unit Maintenance Providers
Environment ■ Clear passage in hallway ■ Floors free of spills ■ Electrical cords secured and not obstructing environment ■ Sharps containers no more than 2/3 full; no item protruding from container ■ Linen room door closed and latched ■ Dirty utility room door closed and latched ■ Supply rooms free of corrugated cardboard ■ Housekeeping cart clean from previous shift
■ ■ ■ ■
Housekeeping closet free of clutter Chemicals properly stored (securely locked and not stored above head level) Unit free of employee food and drink Trash container lids clear of items
Equipment ■ Nonfunctioning equipment labeled; copy of work order attached ■ Wet vacuum available in closet ■ Brakes on wheelchairs properly functioning ■ Refrigerator temperature monitored
* prn, as needed; IV, intravenous; NG, nasogastric; EKG, electrocardiogram; ID, identification.
310
JUNE 2002 JOURNAL Copyright 2002 Joint Commission on Accreditation of Healthcare Organizations
JOURNAL
ON
QUALITY IMPROVEMENT
Table 2. The Three Configurations of the Nursing Safety Standards* ■ ■
■
26 daily standards, divided into a three-part list (Oct 1999–Sep 2000) 4 daily standards + 5 intermittent standards + 5 standards rotating each quarter + 7 ventilator standards (36 total standards; Oct 2000– Jul 2001) 2 daily standards (rotate weekly over 10-week cycle) + 5 intermittent standards (25 total standards; Aug 2001–present)
* Table 1 (p 310) is a comprehensive list of safety standards used at some point during the safety checklist program. Not all of the standards are currently employed.
ventilator complications were added. The ventilator standards included items such as changing endotracheal tube position, providing frequent mouth care, and monitoring sedation levels. The process for monitoring safety standards was further refined in August 2001. Two new standards are posted weekly. Instead of daily forms, a weekly data collection tool is now placed in each room. The five intermittent standards are included on the weekly form. There are a total of 10 unique sheets. Twice daily, at change of shift, the offgoing and oncoming nurses jointly document compliance. The technique is accomplishing several positive practice changes: ■ It removes the opportunity for confirmation bias, defined as the predisposition to “accept information that agrees with our hypothesis and reject information that does not.”15 In the context of safety, this translates to seeing what one expects to see. For example, while replacing tubing on a bank of IVs, a nurse may accidentally reconnect two lines improperly. The nurse may not recognize the error because he or she expects to see the IVs reassembled correctly. However, a colleague may readily unearth the mistake. When the program began, the nurses were asked to check rooms other than their own to avoid the risk of confirmation error. Having two nurses perform the task in tandem greatly reduces the possibility of following the wrong procedure. ■ It serves as an additional impetus for bedside rounds. ■ It essentially eliminates any time burden associated with earlier versions of the checklist. The nurses incorporate the safety checks into their change of shift report, recording compliance information at that time.
VOLUME
28
NUMBER
Weekly rotation of standards keeps the program fresh while allowing each standard to be measured during several weeks interspersed throughout the year. ■ Reducing the number of items being monitored increases compliance with twice-daily checks. ■
Data Management Quick turnaround of information is a key to the program’s success. The program not only provides timely feedback to staff but also indicates that the performance improvement service is a responsive partner in this venture. To accomplish timely posting of results, students from a nearby university were hired on a part-time basis to set up spreadsheets, transfer data from the forms completed on the units to the spreadsheets, and create three visual displays that are exhibited in each ICU. These displays are regularly updated. Figure 2 (p 312) presents a sample bar graph that demonstrates the percentage of compliance with each standard. Respiratory therapists and unit maintenance providers each maintain an established set of 14 safety standards. Although the two groups look at different safety elements specific to the responsibilities of the service, the standards do not change from day to day. The data on percentage of compliance with individual standards for respiratory therapists and unit maintenance providers is summarized and posted monthly. As described earlier, nurses look at 2 new standards each week, with 20 standards measured during a 10-week period. The results are posted at the conclusion of each 10-week cycle. Figure 3 (p 312) presents a second sample bar graph that depicts the number of times each standard is unmet. Because compliance is generally high, aggregate data on the number of incidents of noncompliance for respiratory therapists and unit maintenance providers are posted quarterly rather than monthly. For nurses these data are updated and displayed for each 10-week cycle of standard measurement. Finally, Table 3 (p 313) provides a representative sample of individual incidents of noncompliance with standards. Tables and graphs are posted in the ICU for nurses, respiratory therapists, and unit maintenance providers. Reviewed in combination, the displays permit staff to gather both general and detailed information on the results of their checklist measurements.
6
Copyright 2002 Joint Commission on Accreditation of Healthcare Organizations
311
THE JOINT COMMISSION Representative Sample of Compliance with Individual Safety Standards, Jan–Jun 2001 Registered Nurses’ Compliance with Safety Standards for Ventilated Patients in the Thoracic Intensive Care Unit 100%
100%
100%
100%
% Compliance with Standards
90% 89%
93%
77%
70% 60%
89%
88%
80%
67%
64%
Jan ’01 Feb ’01 Mar ’01 Apr ’01 May ’01 Jun ’01
58%
50% 40% 30%
29%
20% 10% 0% Ramsey Scale Documented Mouth Care Done Ventilated Patient Standards
Figure 2. This representative sample of a bar graph depicts the percentage of compliance with individual standards. Graphs like this one are posted on each unit.
Representative Sample of a Number of Incidents of Noncompliance Noncompliance with Standards for Respiratory Therapists in the Intensive Care Unit
Oxygen equipment changed within 72 hours
12
Ambu bag in room
4
Pulmonary/oral secretion tubing and canister changed within 48 hours
7
5
0
7
1st quarter 2001 2nd quarter 2001 3rd quarter 2001
10
10
5
4
10
4
15
20
25
Number of Incidents of Noncompliance Figure 3. This representative sample of a bar graph depicts the number of incidents of noncompliance with individual safety standards. Graphs
like this one are posted on each unit.
312
JUNE 2002 JOURNAL Copyright 2002 Joint Commission on Accreditation of Healthcare Organizations
JOURNAL
ON
QUALITY IMPROVEMENT
Table 3. Safety Tracking for Registered Nurses in the Thoracic Intensive Care Unit* Date
Room
Topic
Shift
Issue
10/13/01
6A
M
nights
Double signature missing on medication administration record (indicates double-check of high-risk IV infusions)
10/13/01
6B
O
days
Mouth care not documented every 4 hours for ventilated patient
10/16/01
6F
O
nights
Mouth care not documented every 4 hours for ventilated patient
10/18/01
6A
O
days
Pain scale not documented every 4 hours
10/21/01
6E
IV
days
IV pumps and/or tubing not labeled
10/22/01
6C
O
days
Pain scale not documented every 4 hours
10/26/01
6B
O
nights
Pain scale not documented every 4 hours
10/27/01
6F
V
nights
Sedation scale not recorded every 4 hours for ventilated patient
10/27/01
6C
V
nights
Sedation scale not recorded every 4 hours for ventilated patient
10/30/01
6D
O
days
Pain scale not documented every 4 hours
11/1/01
6B
O
nights
Pain scale not documented every 4 hours
11/1/01
6A
V
days
IV pumps and/or tubing not labeled
11/2/01
6D
M
days
Documentation of prn medications not completed
11/5/01
6D
M
nights
Documentation of prn medications not completed
* Dates and room numbers have been changed to protect confidentiality. M, medication; IV, intravenous; O, other; V, ventilator; prn, as needed.
Outcomes Since the inception of the checklist program, thousands of measurements have been taken that reflect either compliance or noncompliance with the safety standards. It is challenging to quantify the safety advances because the checklist configuration has been altered several times, compliance within each discipline (for example, nurses, respiratory therapists, unit maintenance personnel) varies by individual ICU, and even within each ICU compliance with several standards has varied over time. Moreover, a few of the standards were originally measured by one discipline but are currently measured by another group. For example, nurses initially collected data on the dating and replacement of pulmonary suction tubing and canisters. Now respiratory therapists gather that information. Improved compliance with these select standards, however, does not address the truly critical outcomes. These outcomes may be best pursued by seeking the answers to five questions: 1. Have there been recurrences of the same type of medication error, involving nonsecure medications on carts, which led to the original sentinel event? 2. Have there been similar medication errors?
VOLUME
28
NUMBER
3. Have patient deaths or cases of permanent morbidity linked to safety deficits occurred since the inception of the program? 4. Have staff been able to transfer their learning and experience about a specific problem to the general concept of creating a safe environment instead of solely focusing on discrete safety elements? 5. Do staff attitudes regarding the occurrence of error differ as compared with before the program’s inception? There have been no recurrences of error linked to improper storage of medications on top of carts. However, in the past 2 years there have been two errors with a significant risk of morbidity or mortality. Neither incident resulted in permanent injury. The first error occurred during the early months of program implementation. IV potassium was infused through an arterial line for approximately 8 hours. One nurse improperly connected the line just before midnight, but the nurse caring for the patient during the following shift failed to perform the prescribed safety check. A third nurse who entered duty at 8:00 AM discovered the improper connection, which alerted the staff to the importance of the checklist. The second incident took place recently and involved hanging via IV piggyback (IVPB) a neuromuscular blocker on
6
Copyright 2002 Joint Commission on Accreditation of Healthcare Organizations
313
THE JOINT COMMISSION a patient being weaned from the ventilator. The IVPB was intended for another patient and was properly labeled. A major proximal cause of the error was unsafe separation and storage of IVPB medications before administration. Although in retrospect the risky practice was transparent, before this event the staff had not associated the safety concerns connected with multidose vial handling with other forms of medication administration. The transfer of the learning experience occurred after the event and is an area where there is room for continued growth. The fourth question deals with staff attitudes toward safety and the incorporation of the “big safety picture” into daily practice. Indirect measures indicate that the program is achieving that goal. Staff are invested in the initiative, as evidenced by their ongoing revisions of both the safety standards and the methods of data collection. The clinical risk manager regularly queries staff about the value of the program. Nurses, respiratory therapists, and unit maintenance providers are able to define a culture of safety and describe the positive impacts that the program has had on unit-level safety. Reaction to the recent IVPB error reflects a change in attitude since the time of the initial sentinel event. The nurse involved immediately disclosed the error to her supervisor and was well supported. Systems weaknesses were reviewed that day, and several changes were made, to reduce the likelihood of recurrence. Compliance with those changes is being monitored through the checklist program and will remain a daily safety standard until the practices are deeply ingrained in the routines of pharmacists and nurses. This response was a major positive shift from the initial reaction to the original sentinel event that engendered the safety checklist. It speaks eloquently of the growth that has transpired in this staff. Occasionally there is slippage in compliance with select safety standards. Posting the data and bringing weaknesses to the attention of the staff has corrected most difficulties. Supervisory support is critical. Unit leaders must point out the problems without blaming individuals or pitting one shift against another. The leaders must also guide staff in formulating solutions.
Lessons Learned Initially, many staff members believed that this project was a
Staff trust is critical to program success.
314
means of challenging and/or measuring their competence. They saw the vocalized intent of using the checklist to improve patient safety as a thinly veiled disguise for assessing individual performance. This deeply rooted conviction was often not openly expressed but nonetheless limited early participation. It persisted for a time, despite efforts at education, including viewing the video Beyond Blame16 and group discussions of a systems approach to medical error. The facility had a cultural history that was often quick to blame individuals deemed responsible for error. It was naïve to think that this culture could change within weeks. Staff observation of leadership behaviors was critical. Over the course of several months, clinicians began to trust leadership actions. After a shaky start-up, nurses, respiratory therapists, and unit maintenance personnel began to understand that less-than-perfect performance did not reflect negatively on professional practice. Partnerships create vital programs. The safety checklist program is a team effort between staff, unitlevel leadership, the chief of staff, and the clinical risk manager. These partners recognize their interdependence and celebrate their accomplishments. Team members acknowledge that they need to remain faithful to their portion of the agreement. Nurses and ancillary staff must complete the safety checks, and the clinical risk manger is responsible for technical support and rapid turnaround of aggregate data. Maintaining the program takes effort. The initial excitement, concerns, and adjustments to the checklist have faded, but there is a need to keep the program going over the years. Unless carefully nurtured, the momentum will diminish. To provide ongoing support, the clinical risk manager regularly visits the ICUs and listens to staff, both one-on-one and in small groups. During intermittent group meetings, checklist data are shared and the group discusses how to enhance the effort. Finally, clinicians remain excited if leadership remains committed and enthusiastic.
Future Directions Now that the ICU safety checklist program is well established, discussions have been held on how to enhance this initiative. The program will be widened to include input from patients and families.17–18 Stakeholders will be informed about the initiative through
JUNE 2002 JOURNAL Copyright 2002 Joint Commission on Accreditation of Healthcare Organizations
JOURNAL
ON
QUALITY IMPROVEMENT
a brochure. A focus group composed of several patients and family members, staff clinicians, and the clinical risk manager will be formed to elicit ideas for consumer involvement. The second thrust is to expand the program throughout the organization. Giving the ICU staff an opportunity to serve as expert consultants to other areas within the organization that are attempting to develop their own safety programs will give the pioneers much-deserved recognition and also ensure their continued enthusiasm for their own program. A tentative schedule for spread has been established, beginning with mental health and the operating room. Lessons learned through these initial years should smooth introduction of the process, thus reducing start-up time.
Summary Twice-daily safety checks performed by staff in direct care roles are critical in maintaining a constant level of safety. Although the safety standards need to be sound and reflect aspects of care that affect safety, achieving
the “perfect” set of standards is not the core goal. The critical objective is to have clinicians involved in hands-on care incorporate a culture of safety into their daily practice; it is to stand back from the discrete details of their complex work and contemplate the broader picture of how they can improve safety. The project has several key advantages: ■ Bedside clinicians are responsible for and invested in developing safety standards and measuring their practice against those standards; ■ A solid partnership is forged between clinicians and the clinical risk manager, a vital component for effective safety efforts; and ■ Top organizational leaders are provided an opportunity to demonstrate their promise of creating a culture beyond blame. Although not measured in this program, the potential for untoward events, including medication and IV errors, nosocomial infections, ventilator complications, and restraint complications, may be reduced through maintaining a unit-level safety checklist. J
References 1. Institute of Medicine: To Err Is Human: Building a Safer Health System. Washington, DC: National Academy Press, 2000. 2. Institute of Medicine: Crossing the Quality Chasm: A New Health System for the 21st Century. Washington, DC: National Academy Press, 2001. 3. Loeb J: To err is human: An interview with the Institute of Medicine’s Linda Kohn. Jt Comm J Qual Improv 26:227–234, 2000. 4. O’Leary D: Organizational evaluation and a culture of safety. In Proceedings of Enhancing Patient Safety and Reducing Errors in Health Care, Nov 8–10, 1998. Chicago: National Patient Safety Foundation, American Medical Association, 1999, pp 34–37. 5. Berman S: The AMA clinical quality improvement forum on addressing patient safety. Jt Comm J Qual Improv 26:428–433, 2000.
VOLUME
28
NUMBER
6. Phillips DF: “New Look” reflects changing style of patient safety enhancement. JAMA 281: 217–219, 1999. 7. Leape LL (moderator): Creating a culture of safety. In Proceedings of Enhancing Patient Safety and Reducing Errors in Health Care, Nov 8–10, 1998. Chicago: National Patient Safety Foundation, American Medical Association, 1999, pp 23–24. 8. Donchin Y, et al: A look into the nature and causes of human errors in the intensive care unit. Crit Care Med 23:294–300, 1995. 9. Reinertsen JL: Network management and a culture of safety. In Proceedings of Enhancing Patient Safety and Reducing Errors in Health Care, Nov 8–10, 1998. Chicago: National Patient Safety Foundation, American Medical Association, 1999, pp 28–30.
10. Kizer K: Large system change and a culture of safety. In Proceedings of Enhancing Patient Safety and Reducing Errors in Health Care, Nov 8–10, 1998. Chicago: National Patient Safety Foundation, American Medical Association, 1999, pp 31–33. 11. Leape LL: Can we make health care safe? In Reducing Medication Errors and Improving Patient Safety: Success Stories from the Front Lines of Medicine (report). Accelerating Change Today (conference). National Coalition on Health Care (Washington, DC) and the Institute for Healthcare Improvement (Boston), Feb 2000, pp 2–3. 12. Giraud T, et al: Iatrogenic complications in adult intensive care units: A prospective two-center study. Crit Care Med 21:40–51, 1993. 13. Rubins HB, Moskowitz MA: Complications of care in a medical
intensive care unit. J Gen Intern Med 5:104–109, 1990. 14. Hupcey JE: Feeling safe: The psychosocial needs of ICU patients. J Nurs Sch 32:361–367, 2000. 15. Cohen M. Reducing medication errors: We already know what to do! (presentation at Veterans Health Administration and Institute for Healthcare Improvement Collaborative on Reducing Adverse Drug Events). Baltimore, Apr 11, 2000. 16. Bridge Medical: Beyond Blame (videotape) documentary. Solana Beach, CA, 1997. 17. Turning medical error tragedy into opportunity. Executive Solutions for Healthcare Management 3:9–11, Jan 2000. 18. Want a savvy participant in your error-prevention program? Put a consumer on your team! ISMP Medication Safety Alert 5(10):1, 2000.
6
Copyright 2002 Joint Commission on Accreditation of Healthcare Organizations
315