- Email: [email protected]
Helping Hippocrates: A Cross-Functional Approach to Patient Identification
National Patient Safety Goals
Helping Hippocrates: A Cross-Functional Approach to Patient Identification
Margaret A. Greenly, L.C.S.W., M.P.A. Department Editors: Marcia M. Piotrowski, R.N., M.S., Peter Angood, M.D., Paula Griswold, M.S., Gina Pugliese, R.N., M.S., Sanjay Saint, M.D., M.P.H., Susan E. Sheridan, M.I.M., M.B.A., Kaveh G. Shojania, M.D. Readers may submit National Patient Safety Goals inquiries and submissions to Steven Berman (sberman@jcaho. org) and Marcia Piotrowski (marcia. [email protected]).
Case Study Mrs. C. (not actual initial) was admitted to the “D” (door) bed in room 715. Her physician called in admitting orders, including an order for a colonoscopy. The nurse taking that order followed the read-back procedure after writing the order in the medical record. The unit secretary entered the order into the computer. Mrs. C., as ordered, underwent the prep for the colonoscopy that afternoon and evening. Mrs. C’s roommate was discharged shortly after Mrs. C.’s arrival, and Mrs. C. asked if she could move to the “W” (window) bed. Her nurse agreed, printed a new identification (ID) band and changed the cover of the medical record to reflect the new location. A new patient, Mrs. M. was admitted to the “D” bed. The next morning, a transporter arrived and announced, “I’m here to take 715D down to Endoscopy.” The unit secretary directed him to the room and handed him the medical record for Mrs. M. The nurse (a different nurse from the day before) assisted him in transferring her to a stretcher and she was taken to Endoscopy, placed on the table, and readied for her procedure. Fortunately, the physician had previously seen Mrs. C. in his office, and, upon entering the endoscopy suite, realized immediately that Mrs. M. was the incorrect patient. Her ID band was compared with the order, and the error was discovered. No harm occurred to Mrs. M., and Mrs. C. did have her procedure later that day. When informed of the error, Mrs. M. expressed disappointment that she
August 2006
Article-at-a-Glance Background: The Joint Commission on Accreditation of Healthcare Organizations National Patient Safety Goal 1, which requires the use of at least two patient identifiers, is the foundation for other patient safety goals. St. Francis Hospital involved staff and patients in the “Helping Hippocrates” Project, which used a “game” with staff and patients to ensure the accuracy of information on patients’ identification (ID) bands. The Project: Members of all hospital departments assigned to a specific day were to compare the ID band with the patient census report and identify patients who had no ID band on their wrist and patients who had a band with inaccuracies. They were to also ask patients if the staff had checked the ID band before treatments or procedures. Also, the nurse manager was to select a patient to add to his or her own ID band a special band bearing the name Hippocrates. The department conducting the survey had to find Hippocrates. Findings: Internal data showed that patient identification errors declined from 8.2% to a sustained zero. Patient satisfaction data showed that since the inception of Helping Hippocrates, patients’ perceptions of staff’s compliance with ID verification showed steady improvement. Conclusion: Helping Hippocrates demonstrates the value of using an innovative problem-solving strategy that engages the entire organization.
Volume 32 Number 8
Copyright 2006 Joint Commission on Accreditation of Healthcare Organizations
463
did not have the colonoscopy, saying, “I figured it would be a good idea to get it out of the way,” as her doctor had discussed its importance at her last annual physical. A review of this event identified the following breakdowns in the patient ID process: 1. The transporter asked for the patient by location. 2. The unit secretary gave him a medical record, without verifying the name of the patient. 3. The nurse at the bedside did not verify the patient’s ID against the order before transferring her to the stretcher. 4. The nurse in endoscopy also did not verify ID against the order. Process changes then instituted included the requirement that transporters ask for patients by name, not room number. The unit secretary is then to verify the name and medical record number, before handing the medical record to the transporter. Nurses have all been reeducated on their responsibility to verify the patient’s ID before allowing a patient to be transported to a procedure. Nurses and technologists in all procedure areas (not just endoscopy) have also been reeducated that they must verify the patient’s ID before placing a patient on the table.
example, O’Neil described the “close call” that occurred due to an error in patient identification in a pediatric emergency room.3 Memorial Sloan-Kettering Cancer Center has worked to establish and support a systemwide approach to correct patient identification.4 This article describes a unique approach to ensuring that ID bands are in place and contain the correct information. During a six-month period in 2004, the patient care safety committee (PCSC) at St. Francis Hospital, a community hospital in Poughkeepsie, New York, actively involved staff and patients in the “Helping Hippocrates” Project. The project’s name reflects the Hippocratic Oath, “I will keep them from harm and injustice,” which stems from fifth-century B.C.5 (The nursing oath, written by Florence Nightingale,6 reflects the same commitment to patients’ welfare that nurses—and all health care professionals—accept.) This project used a “game” to raise consciousness and accountability for 100% accuracy of patient ID bands. The Joint Commission recognizes the importance of empowering patients to become active participants in assuring safety, as demonstrated by the “Speak Out” campaign for patient education and action regarding safety.7
Background
The Project
The Joint Commission on Accreditation of Healthcare Organizations National Patient Safety Goal 1, in effect since January 2003, requires that health care organizations use at least two patient identifiers (neither to be the patient’s room number) whenever administering medications or blood products; taking blood samples and other specimens for clinical testing, or providing any other treatments or procedures.1 This first goal is the foundation for other patient safety goals—for to do anything else correctly we must first be sure we have the correct patient. Patient ID bands are not a requirement for meeting this goal, and, alone, do not satisfy the requirement. It is the person-specific information that is the identifier, not the medium on which that information is located.2 To be useful as a medium, the patient ID band must be on the patient and must contain accurate information for the staff to use as a reference. Otherwise, the risk of error is clearly present. The large number of articles on patient identification in the literature indicate the importance of this issue. For
Origins in Experience and Data
464
August 2006
In 2003, utilizing the incident reporting system, laboratory staff reported anecdotal information that identified a major issue concerning patient identification: At times, it was not possible for phlebotomists to verify patient identification before drawing blood because patients either had no ID bands on at all, or had bands with incorrect or missing information. In October 2003, we (PCSC) formalized the laboratory’s anecdotal information into an indicator and began to measure the incidence of such events. We created a policy that stated, “Everyone is responsible for patient identification.” However, the laboratory continued to identify such events each month, sometimes as few as one, but reaching an alarming 18 in January 2004. The sentinel event review committee has reviewed no actual sentinel events related to patient identification; however, it has reviewed several close calls. These included the incorrect patient being brought to a test or procedure, a potential medication error, and a situation
Volume 32 Number 8
Copyright 2006 Joint Commission on Accreditation of Healthcare Organizations
in which patients with similar names had a mix-up in ID bands. Although all were caught in time, and no harm resulted to any of the patients, these were clear indications of a systems breakdown.
The Need for a Unique Approach On the basis of information from the laboratory and the close calls reviewed by the sentinel event review committee, the PCSC recognized the risk for a sentinel event caused by weaknesses in the patient identification process. It decided to adopt correct patient identification as its major focus. A small team was formed in February 2004, consisting of two physicians (one from the emergency department and one from mental health), two nurse managers (one from a medical-surgical floor and one from mental health), a pharmacist, and the director of quality improvement [M.A.G.] The team was charged with implementing a unique method to strengthen the patient identification process throughout the hospital—two medical-surgical units, three critical care units, three mental health units, and an acute rehabilitation unit. The team began its work by interviewing staff from across the organization—and patients—to better assess the culture within which this problem was occurring. Its key findings were as follows: ■ There was a perception that, if nurses verified the information on the ID band at the start of every shift, other staff could be confident that the information was accurate when they then used the ID band as one of the sources of the two identifiers. Nurses pointed out the fallacy of this perception. For example, an ID band may be on at the start of a shift but be off 10 minutes later, especially for mental health patients and elderly and confused patients. Also, they were concerned that giving the responsibility to one discipline could lead others (for example, radiology, laboratory) to assume that it is correct and therefore not have “to bother” checking it themselves. In addition, nurses admitted to feeling defensive, saying they are “responsible for everything” when, in fact, multiple people touch the process of patient care. ■ When laboratory staff members pointed out to a nurse that a patient was not wearing a band, the nurse sometimes reacted in a way that appeared
August 2006
hostile but was actually an issue of pride: “I take good care of my patients and I know each of them personally. If I tell you the man in that bed is Mr. X, then you can believe me. “ ■ In spite of the policy of verifying the two identifiers before any intervention, some staff members, such as transporters, still referred to patients by their room numbers. ■ Some members of other departments who provided treatment did not check the ID band because they did not know the policy applied to them. One physical therapist admitted, “Honestly, it never occurred to me. I just assumed someone else had done it. “ ■ Some staff members uninvolved in the patient ID process, such as employees in the billing and payroll offices, who were unfamiliar with the complexities of caring for patients while juggling multiple priorities, did not understand the problem. As one employee stated, “Well, what’s so hard about it? Why can’t the nurses just do it?” ■ Patients were very clear that verifying their ID was extremely important to them. Some patients voiced their perception that adherence to this process was inconsistent. (“The nurses work hard…maybe they’re too busy,” “The nurse checks when she first comes on duty, but not always after that,” “The nurses check, but not other people.”)
Philosophical Assumptions In accordance with its stated organizational commitment to create and maintain an error-free and blame-free environment, the team established core assumptions: ■ The foundation of patient safety is correct patient identification. For this to happen, patients must, at all times, wear accurate ID bands. ■ Every employee, clinical or nonclinical, must follow the National Patient Safety Goal requirements. ■ A written policy alone is not a solution. ■ We cannot assume that “someone” else has verified the patient’s identification. ■ The patient identification process requires us to work together, beyond the boundaries of our individual departments or areas (outside our “silos”). ■ Nagging is not an effective intervention. ■ Patients, as well as all staff, have wisdom to offer.
Volume 32 Number 8
Copyright 2006 Joint Commission on Accreditation of Healthcare Organizations
465
Creating and Implementing the Helping Hippocrates Approach The team created the Helping Hippocrates approach to address all the elements of its charge from the PCSC. The approach was designed to be cross-functional, including even departments not involved in patient care, and patients themselves—and to help “create joy in work,”8 as Deming advocated. We created a rotational assignment for weekly verification of patient ID. This avoided any one department’s having to bear the complete responsibility. The assignment included all seven nursing units/departments, as well as all other departments, such as radiology; laboratory; respiratory therapy; electrocardiography (EKG); physical and occupational therapy; surgical services; and even the billing office, medical records, housekeeping, and plant operations. Each department was to be assigned a day during which they were to visit a specific nursing unit and conduct a bedside survey of all patients on that unit. Using a data collection tool designed by the team, members of the assigned department were to compare the ID band with the patient census report. They were to identify patients who had no ID band on their wrist and patients who had a band with inaccuracies, such as an incorrect or misspelled name or an incorrect medical record number. When appropriate, they were to ask individual patients if the staff had checked the ID band before treatments or procedures and recorded the patients’ replies in their own words. The data collection sheets were to be sent to the quality improvement services department for aggregation and analysis. Department members were to be assigned to locations they would not usually frequent, which enhanced familiarity with the patient care responsibilities of others (for example, the operating room staff to adolescent mental health, adolescent mental health to critical care, and respiratory therapy to rehabilitation). Departments that were not clinical, such as billing, were to be randomly assigned to any patient care unit. The nurse manager of each unit was to know in advance what day his or her unit would be surveyed. However, the staff were not to know, to prevent advance preparation for the survey. Shortly before the survey team was to arrive on the unit, the nurse manager was to select a patient and ask
466
August 2006
him or her to help us “play a game” with the staff. The “game” was that the patient was to be given, in addition to his or her own ID band, a special band, bearing the name Hippocrates. The department conducting the survey had to find Hippocrates. After finding Hippocrates and after turning in their data collection sheets on all patients on the unit, department staff would receive a tray of cookies as thanks. We had concerns about how patients would respond to the Helping Hippocrates Approach, so we conducted a pilot test. We found that they were very pleased about the effort we were making to ensure they were correctly identified. The patients who played the role of Hippocrates in the pilot were uniformly delighted with the special attention and the idea of “playing a trick” on staff. We found some flaws in our data collection tool and revised it. For example, we found that we needed to spell out the instructions very clearly because a wide range of staff, including those not accustomed to patient contact, would be using the tool. Also, we added a “comments” column because patients sometimes brought up other issues, unrelated to the survey, that we wanted to be sure were followed up on. We were then ready to fully implement the program. The team kicked off the campaign at a leadership meeting. To set the tone, we switched ID badges with one another or with other members of management. For example, one team member who is a female nurse wore the ID badge of the male chief executive officer. Although we positioned ourselves at every doorway into the conference room, and greeted each manager by name and with a handshake, very few people noticed our incorrect badges. At the start of our presentation, we used this to point out that we tend to see what we expect to see—and that tendency is a major cause of errors. The meeting then became serious. Using a slide presentation, the team described the risks associated with incorrect patient identification. We described several cases from the news media in which patients had been harmed or died as a result of such errors as well as the close calls reviewed by the sentinel event review committee. We posed the mother-standard: “If this was your mother, what would you want?” Staff were informed about the project at staff meetings and an article in The Grapevine, the widely read
Volume 32 Number 8
Copyright 2006 Joint Commission on Accreditation of Healthcare Organizations
weekly hospital newsletter. The target date for implementation was set for two weeks following this publication. We fully implemented the project in March 2004 and continued it consistently during a six-month period, with random checks since then.
Percentage of Patient Identification (ID) Bands Missing or Inconsistent on Survey
Findings The data collection sheets were collated as they arrived in the quality improvement services department, and the findings were aggregated every three weeks. As shown in Figure 1 (right), the Figure 1. The initial error rate was 8.2% but steadily decreased during the duration of the six-month project (T-1–T-8) until it reached zero—remaininitial error rate was 8.2% but steadily ing at that level for 15 months. The times noted reflect data aggregated decreased during the duration of the every three weeks from the weekly surveys. project until it reached zero—remaining which a noncompliance rate of 1% was documented. at that level for 15 months. Two patients (on the same unit) stated that they had Department members conducting the survey were to been there awhile and that most of the nurses know ask patients if the staff checked their ID band before them and that only the new nurses check. Immediate any medication or procedure. Some patients were too ill reeducation was provided to the staff on that unit, with or too confused to be asked. Those who responded gave subsequent complete compliance. a variety of answers, ranging from “Yes, always” to Another source of data has been the laboratory “Well, the nurses usually do, but not everyone else,” to report on patient ID errors identified by phlebotomists. “I guess so,” to “No.” Any response other than “Yes, As shown in Figure 3 (page 468), the number of always,” was scored as negative. As shown in Figure 2 patients with ID band errors also demonstrated a (below), we achieved a steady decline in the frequency of negative responses to zero—until one survey time in steady and sustained decline, from the alarming 26 in the first quarter of 2004 to a sustained zero for five quarters. A third source of data was our patient satisfaction survey. We participate in the Picker survey and had two comparative groups available—the other participating hospitals in NorMet (the Northern Metropolitan Region of New York State, that is, counties between New York City and Albany) and other 200–300-bed hospitals from across the United States. A randomly selected sample of patients from an electronic file submitted weekly to Picker were sent the survey a few weeks following discharge. One quesFigure 2. The percentage of patients providing negative replies steadily decreased during the duration of the six-month project (T-1–T-8). tion on the survey asked, “Did the
Percentage of Negative Responses to Whether Staff Check Identification (ID) Bands
August 2006
Volume 32 Number 8
Copyright 2006 Joint Commission on Accreditation of Healthcare Organizations
467
Patients with Identification (ID) Band Errors Identified by Phlebotomists
chair, with her bags packed and ready for discharge, declared, “My doctor says I’m ready to go home, but I told my son he couldn’t come here to pick me up until you found me!”
Conclusion “Helping Hippocrates” demonstrates the value of using an innovative problemsolving strategy that engages the entire organization, including patients themselves, and has rewards built into it. The principles learned through “Helping Hippocrates” can be of enduring value to our organization. Figure 3. The number of patients with ID band errors demonstrated a Data from three different sources valsteady and sustained decline. The Helping Hippocrates project was impleidated the success of the project. From mented from March through August 2004. an interpersonal and cultural point of nurses check your ID band before giving you medicaview, it accomplished even more. It took the burden off tions or starting a procedure?” As shown in Figure 4 nursing and helped clinical staff appreciate the scope of (below), in the first quarter of 2004, our score on this one another’s work as they visited different units to conquestion was statistically significantly lower than that duct the surveys. For nonclinical departments, it was an of the two comparison groups. Yet our performance on eye-opening experience. In the words of one staff member, who had never been to an inpatient unit, “Now I see this steadily improved since the inception of “Helping why my job in Finance matters!” Hippocrates” in the second quarter of 2004, and the improvement has been sustained through the most recent report (fourth quarter of 2005). Other findings were anecdotal. Staff who conducted the surveys took their responsibility seriously and enjoyed the “game” approach and the reward of the cookies. We found that, as expected, no single department bore an inordinate share of the work associated with this project. The patients who played the role of Hippocrates during the project unanimously supported it. They said that they felt glad to be part of something that would make all patients safer. Several added their own twist of humor, includFigure 4. St. Francis Hospital (SFH)’s performance on this question ing giving the Hippocrates ID band as a steadily improved since the “Helping Hippocrates” project (T-1–T-8) dursouvenir to the team that found them. ing the second quarter of 2004, and the improvement has been sustained through the fourth quarter of 2005. One “Hippocrates,” who was sitting in a
Patients’ Postdischarge Perception of Patient Identification (ID) Compliance
468
August 2006
Volume 32 Number 8
Copyright 2006 Joint Commission on Accreditation of Healthcare Organizations
Postscript St. Francis Hospital’s “Helping Hippocrates” Project was submitted to the New York State Department of Health’s 2005 Patient Safety Award. We are honored to have been the only award recipient in our category (200+ beds).9 Our $200,000 award will be used to help implement bar-coding medication administration. Our challenge now will be to ensure that all caregivers do not become complacent and
assume that bar coding has eliminated their responsibility to constantly and consistently verify patient IDs. J
Margaret A. Greenly, L.C.S.W., M.P.A., is Director, Quality Improvement and Case Management, St. Francis Hospital, Poughkeepsie, New York. Please send reprint requests to Margaret A. Greenly, [email protected].
References 1. Joint Commission on Accreditation of Healthcare Organizations: Facts About National Patient Safety Goals. http://www.jointcommission.org/ PatientSafety/NationalPatientSafetyGoals/06_npsg_facts.htm (last accessed Jun. 1, 2006). 2. Joint Commission on Accreditation of Healthcare Organizations: FAQs for the 2006 National Patient Safety Goals. http://www.joint commission.org/NR/rdonlyres/2E5C5B8C-108D-4A599303-8605B5EB3B0D/0/06_npsg_faq1.pdf (last accessed Jun. 1, 2006). 3. O’Neil K., et al.: Patient misidentification in a pediatric emergency department: Patient safety and legal perspectives. Pediatr Emerg Care 20:487–492, Jul. 2004. 4. Parisi L.: Patient identification: The foundation for a culture of patient safety. J Nurs Care Qual 18:73–79, Jan.–Mar. 2003.
August 2006
5. University of Virginia Health System: Hippocrates. http://www.health system.virginia.edu/internet/library/historical/artifacts/antiqua/ hippocrates.cfm (last accessed Jun. 1, 2006). 6. CyberNurse: Florence Nightingale Pledge. http://www.cybernurse.com/ florencepledge.html (last accessed Jun. 2, 2006). 7. Joint Commission on Accreditation of Healthcare Organizations: Speak Up Initiatives. http://www.jointcommission.org/PatientSafety/ SpeakUp/ (last accessed Jun. 1, 2006). 8. Personal Success and Leadership Institute: Joy in Work. http://pslinstitute.com/joyinwork.html (last accessed Jun. 2, 2006). 9. New York State Department of Health: Dear Administrator Letter Regarding 2006 Patient Safety Awards. http://www.health.state.ny.us/ professionals/patients/patient_safety/awards/2006_dhal_pat_safety_a ward.htm (last accessed Jun. 2, 2006).
Volume 32 Number 8
Copyright 2006 Joint Commission on Accreditation of Healthcare Organizations
469
Helping Hippocrates: A Cross-Functional Approach to Patient Identification
Margaret A. Greenly, L.C.S.W., M.P.A. Department Editors: Marcia M. Piotrowski, R.N., M.S., Peter Angood, M.D., Paula Griswold, M.S., Gina Pugliese, R.N., M.S., Sanjay Saint, M.D., M.P.H., Susan E. Sheridan, M.I.M., M.B.A., Kaveh G. Shojania, M.D. Readers may submit National Patient Safety Goals inquiries and submissions to Steven Berman (sberman@jcaho. org) and Marcia Piotrowski (marcia. [email protected]).
Case Study Mrs. C. (not actual initial) was admitted to the “D” (door) bed in room 715. Her physician called in admitting orders, including an order for a colonoscopy. The nurse taking that order followed the read-back procedure after writing the order in the medical record. The unit secretary entered the order into the computer. Mrs. C., as ordered, underwent the prep for the colonoscopy that afternoon and evening. Mrs. C’s roommate was discharged shortly after Mrs. C.’s arrival, and Mrs. C. asked if she could move to the “W” (window) bed. Her nurse agreed, printed a new identification (ID) band and changed the cover of the medical record to reflect the new location. A new patient, Mrs. M. was admitted to the “D” bed. The next morning, a transporter arrived and announced, “I’m here to take 715D down to Endoscopy.” The unit secretary directed him to the room and handed him the medical record for Mrs. M. The nurse (a different nurse from the day before) assisted him in transferring her to a stretcher and she was taken to Endoscopy, placed on the table, and readied for her procedure. Fortunately, the physician had previously seen Mrs. C. in his office, and, upon entering the endoscopy suite, realized immediately that Mrs. M. was the incorrect patient. Her ID band was compared with the order, and the error was discovered. No harm occurred to Mrs. M., and Mrs. C. did have her procedure later that day. When informed of the error, Mrs. M. expressed disappointment that she
August 2006
Article-at-a-Glance Background: The Joint Commission on Accreditation of Healthcare Organizations National Patient Safety Goal 1, which requires the use of at least two patient identifiers, is the foundation for other patient safety goals. St. Francis Hospital involved staff and patients in the “Helping Hippocrates” Project, which used a “game” with staff and patients to ensure the accuracy of information on patients’ identification (ID) bands. The Project: Members of all hospital departments assigned to a specific day were to compare the ID band with the patient census report and identify patients who had no ID band on their wrist and patients who had a band with inaccuracies. They were to also ask patients if the staff had checked the ID band before treatments or procedures. Also, the nurse manager was to select a patient to add to his or her own ID band a special band bearing the name Hippocrates. The department conducting the survey had to find Hippocrates. Findings: Internal data showed that patient identification errors declined from 8.2% to a sustained zero. Patient satisfaction data showed that since the inception of Helping Hippocrates, patients’ perceptions of staff’s compliance with ID verification showed steady improvement. Conclusion: Helping Hippocrates demonstrates the value of using an innovative problem-solving strategy that engages the entire organization.
Volume 32 Number 8
Copyright 2006 Joint Commission on Accreditation of Healthcare Organizations
463
did not have the colonoscopy, saying, “I figured it would be a good idea to get it out of the way,” as her doctor had discussed its importance at her last annual physical. A review of this event identified the following breakdowns in the patient ID process: 1. The transporter asked for the patient by location. 2. The unit secretary gave him a medical record, without verifying the name of the patient. 3. The nurse at the bedside did not verify the patient’s ID against the order before transferring her to the stretcher. 4. The nurse in endoscopy also did not verify ID against the order. Process changes then instituted included the requirement that transporters ask for patients by name, not room number. The unit secretary is then to verify the name and medical record number, before handing the medical record to the transporter. Nurses have all been reeducated on their responsibility to verify the patient’s ID before allowing a patient to be transported to a procedure. Nurses and technologists in all procedure areas (not just endoscopy) have also been reeducated that they must verify the patient’s ID before placing a patient on the table.
example, O’Neil described the “close call” that occurred due to an error in patient identification in a pediatric emergency room.3 Memorial Sloan-Kettering Cancer Center has worked to establish and support a systemwide approach to correct patient identification.4 This article describes a unique approach to ensuring that ID bands are in place and contain the correct information. During a six-month period in 2004, the patient care safety committee (PCSC) at St. Francis Hospital, a community hospital in Poughkeepsie, New York, actively involved staff and patients in the “Helping Hippocrates” Project. The project’s name reflects the Hippocratic Oath, “I will keep them from harm and injustice,” which stems from fifth-century B.C.5 (The nursing oath, written by Florence Nightingale,6 reflects the same commitment to patients’ welfare that nurses—and all health care professionals—accept.) This project used a “game” to raise consciousness and accountability for 100% accuracy of patient ID bands. The Joint Commission recognizes the importance of empowering patients to become active participants in assuring safety, as demonstrated by the “Speak Out” campaign for patient education and action regarding safety.7
Background
The Project
The Joint Commission on Accreditation of Healthcare Organizations National Patient Safety Goal 1, in effect since January 2003, requires that health care organizations use at least two patient identifiers (neither to be the patient’s room number) whenever administering medications or blood products; taking blood samples and other specimens for clinical testing, or providing any other treatments or procedures.1 This first goal is the foundation for other patient safety goals—for to do anything else correctly we must first be sure we have the correct patient. Patient ID bands are not a requirement for meeting this goal, and, alone, do not satisfy the requirement. It is the person-specific information that is the identifier, not the medium on which that information is located.2 To be useful as a medium, the patient ID band must be on the patient and must contain accurate information for the staff to use as a reference. Otherwise, the risk of error is clearly present. The large number of articles on patient identification in the literature indicate the importance of this issue. For
Origins in Experience and Data
464
August 2006
In 2003, utilizing the incident reporting system, laboratory staff reported anecdotal information that identified a major issue concerning patient identification: At times, it was not possible for phlebotomists to verify patient identification before drawing blood because patients either had no ID bands on at all, or had bands with incorrect or missing information. In October 2003, we (PCSC) formalized the laboratory’s anecdotal information into an indicator and began to measure the incidence of such events. We created a policy that stated, “Everyone is responsible for patient identification.” However, the laboratory continued to identify such events each month, sometimes as few as one, but reaching an alarming 18 in January 2004. The sentinel event review committee has reviewed no actual sentinel events related to patient identification; however, it has reviewed several close calls. These included the incorrect patient being brought to a test or procedure, a potential medication error, and a situation
Volume 32 Number 8
Copyright 2006 Joint Commission on Accreditation of Healthcare Organizations
in which patients with similar names had a mix-up in ID bands. Although all were caught in time, and no harm resulted to any of the patients, these were clear indications of a systems breakdown.
The Need for a Unique Approach On the basis of information from the laboratory and the close calls reviewed by the sentinel event review committee, the PCSC recognized the risk for a sentinel event caused by weaknesses in the patient identification process. It decided to adopt correct patient identification as its major focus. A small team was formed in February 2004, consisting of two physicians (one from the emergency department and one from mental health), two nurse managers (one from a medical-surgical floor and one from mental health), a pharmacist, and the director of quality improvement [M.A.G.] The team was charged with implementing a unique method to strengthen the patient identification process throughout the hospital—two medical-surgical units, three critical care units, three mental health units, and an acute rehabilitation unit. The team began its work by interviewing staff from across the organization—and patients—to better assess the culture within which this problem was occurring. Its key findings were as follows: ■ There was a perception that, if nurses verified the information on the ID band at the start of every shift, other staff could be confident that the information was accurate when they then used the ID band as one of the sources of the two identifiers. Nurses pointed out the fallacy of this perception. For example, an ID band may be on at the start of a shift but be off 10 minutes later, especially for mental health patients and elderly and confused patients. Also, they were concerned that giving the responsibility to one discipline could lead others (for example, radiology, laboratory) to assume that it is correct and therefore not have “to bother” checking it themselves. In addition, nurses admitted to feeling defensive, saying they are “responsible for everything” when, in fact, multiple people touch the process of patient care. ■ When laboratory staff members pointed out to a nurse that a patient was not wearing a band, the nurse sometimes reacted in a way that appeared
August 2006
hostile but was actually an issue of pride: “I take good care of my patients and I know each of them personally. If I tell you the man in that bed is Mr. X, then you can believe me. “ ■ In spite of the policy of verifying the two identifiers before any intervention, some staff members, such as transporters, still referred to patients by their room numbers. ■ Some members of other departments who provided treatment did not check the ID band because they did not know the policy applied to them. One physical therapist admitted, “Honestly, it never occurred to me. I just assumed someone else had done it. “ ■ Some staff members uninvolved in the patient ID process, such as employees in the billing and payroll offices, who were unfamiliar with the complexities of caring for patients while juggling multiple priorities, did not understand the problem. As one employee stated, “Well, what’s so hard about it? Why can’t the nurses just do it?” ■ Patients were very clear that verifying their ID was extremely important to them. Some patients voiced their perception that adherence to this process was inconsistent. (“The nurses work hard…maybe they’re too busy,” “The nurse checks when she first comes on duty, but not always after that,” “The nurses check, but not other people.”)
Philosophical Assumptions In accordance with its stated organizational commitment to create and maintain an error-free and blame-free environment, the team established core assumptions: ■ The foundation of patient safety is correct patient identification. For this to happen, patients must, at all times, wear accurate ID bands. ■ Every employee, clinical or nonclinical, must follow the National Patient Safety Goal requirements. ■ A written policy alone is not a solution. ■ We cannot assume that “someone” else has verified the patient’s identification. ■ The patient identification process requires us to work together, beyond the boundaries of our individual departments or areas (outside our “silos”). ■ Nagging is not an effective intervention. ■ Patients, as well as all staff, have wisdom to offer.
Volume 32 Number 8
Copyright 2006 Joint Commission on Accreditation of Healthcare Organizations
465
Creating and Implementing the Helping Hippocrates Approach The team created the Helping Hippocrates approach to address all the elements of its charge from the PCSC. The approach was designed to be cross-functional, including even departments not involved in patient care, and patients themselves—and to help “create joy in work,”8 as Deming advocated. We created a rotational assignment for weekly verification of patient ID. This avoided any one department’s having to bear the complete responsibility. The assignment included all seven nursing units/departments, as well as all other departments, such as radiology; laboratory; respiratory therapy; electrocardiography (EKG); physical and occupational therapy; surgical services; and even the billing office, medical records, housekeeping, and plant operations. Each department was to be assigned a day during which they were to visit a specific nursing unit and conduct a bedside survey of all patients on that unit. Using a data collection tool designed by the team, members of the assigned department were to compare the ID band with the patient census report. They were to identify patients who had no ID band on their wrist and patients who had a band with inaccuracies, such as an incorrect or misspelled name or an incorrect medical record number. When appropriate, they were to ask individual patients if the staff had checked the ID band before treatments or procedures and recorded the patients’ replies in their own words. The data collection sheets were to be sent to the quality improvement services department for aggregation and analysis. Department members were to be assigned to locations they would not usually frequent, which enhanced familiarity with the patient care responsibilities of others (for example, the operating room staff to adolescent mental health, adolescent mental health to critical care, and respiratory therapy to rehabilitation). Departments that were not clinical, such as billing, were to be randomly assigned to any patient care unit. The nurse manager of each unit was to know in advance what day his or her unit would be surveyed. However, the staff were not to know, to prevent advance preparation for the survey. Shortly before the survey team was to arrive on the unit, the nurse manager was to select a patient and ask
466
August 2006
him or her to help us “play a game” with the staff. The “game” was that the patient was to be given, in addition to his or her own ID band, a special band, bearing the name Hippocrates. The department conducting the survey had to find Hippocrates. After finding Hippocrates and after turning in their data collection sheets on all patients on the unit, department staff would receive a tray of cookies as thanks. We had concerns about how patients would respond to the Helping Hippocrates Approach, so we conducted a pilot test. We found that they were very pleased about the effort we were making to ensure they were correctly identified. The patients who played the role of Hippocrates in the pilot were uniformly delighted with the special attention and the idea of “playing a trick” on staff. We found some flaws in our data collection tool and revised it. For example, we found that we needed to spell out the instructions very clearly because a wide range of staff, including those not accustomed to patient contact, would be using the tool. Also, we added a “comments” column because patients sometimes brought up other issues, unrelated to the survey, that we wanted to be sure were followed up on. We were then ready to fully implement the program. The team kicked off the campaign at a leadership meeting. To set the tone, we switched ID badges with one another or with other members of management. For example, one team member who is a female nurse wore the ID badge of the male chief executive officer. Although we positioned ourselves at every doorway into the conference room, and greeted each manager by name and with a handshake, very few people noticed our incorrect badges. At the start of our presentation, we used this to point out that we tend to see what we expect to see—and that tendency is a major cause of errors. The meeting then became serious. Using a slide presentation, the team described the risks associated with incorrect patient identification. We described several cases from the news media in which patients had been harmed or died as a result of such errors as well as the close calls reviewed by the sentinel event review committee. We posed the mother-standard: “If this was your mother, what would you want?” Staff were informed about the project at staff meetings and an article in The Grapevine, the widely read
Volume 32 Number 8
Copyright 2006 Joint Commission on Accreditation of Healthcare Organizations
weekly hospital newsletter. The target date for implementation was set for two weeks following this publication. We fully implemented the project in March 2004 and continued it consistently during a six-month period, with random checks since then.
Percentage of Patient Identification (ID) Bands Missing or Inconsistent on Survey
Findings The data collection sheets were collated as they arrived in the quality improvement services department, and the findings were aggregated every three weeks. As shown in Figure 1 (right), the Figure 1. The initial error rate was 8.2% but steadily decreased during the duration of the six-month project (T-1–T-8) until it reached zero—remaininitial error rate was 8.2% but steadily ing at that level for 15 months. The times noted reflect data aggregated decreased during the duration of the every three weeks from the weekly surveys. project until it reached zero—remaining which a noncompliance rate of 1% was documented. at that level for 15 months. Two patients (on the same unit) stated that they had Department members conducting the survey were to been there awhile and that most of the nurses know ask patients if the staff checked their ID band before them and that only the new nurses check. Immediate any medication or procedure. Some patients were too ill reeducation was provided to the staff on that unit, with or too confused to be asked. Those who responded gave subsequent complete compliance. a variety of answers, ranging from “Yes, always” to Another source of data has been the laboratory “Well, the nurses usually do, but not everyone else,” to report on patient ID errors identified by phlebotomists. “I guess so,” to “No.” Any response other than “Yes, As shown in Figure 3 (page 468), the number of always,” was scored as negative. As shown in Figure 2 patients with ID band errors also demonstrated a (below), we achieved a steady decline in the frequency of negative responses to zero—until one survey time in steady and sustained decline, from the alarming 26 in the first quarter of 2004 to a sustained zero for five quarters. A third source of data was our patient satisfaction survey. We participate in the Picker survey and had two comparative groups available—the other participating hospitals in NorMet (the Northern Metropolitan Region of New York State, that is, counties between New York City and Albany) and other 200–300-bed hospitals from across the United States. A randomly selected sample of patients from an electronic file submitted weekly to Picker were sent the survey a few weeks following discharge. One quesFigure 2. The percentage of patients providing negative replies steadily decreased during the duration of the six-month project (T-1–T-8). tion on the survey asked, “Did the
Percentage of Negative Responses to Whether Staff Check Identification (ID) Bands
August 2006
Volume 32 Number 8
Copyright 2006 Joint Commission on Accreditation of Healthcare Organizations
467
Patients with Identification (ID) Band Errors Identified by Phlebotomists
chair, with her bags packed and ready for discharge, declared, “My doctor says I’m ready to go home, but I told my son he couldn’t come here to pick me up until you found me!”
Conclusion “Helping Hippocrates” demonstrates the value of using an innovative problemsolving strategy that engages the entire organization, including patients themselves, and has rewards built into it. The principles learned through “Helping Hippocrates” can be of enduring value to our organization. Figure 3. The number of patients with ID band errors demonstrated a Data from three different sources valsteady and sustained decline. The Helping Hippocrates project was impleidated the success of the project. From mented from March through August 2004. an interpersonal and cultural point of nurses check your ID band before giving you medicaview, it accomplished even more. It took the burden off tions or starting a procedure?” As shown in Figure 4 nursing and helped clinical staff appreciate the scope of (below), in the first quarter of 2004, our score on this one another’s work as they visited different units to conquestion was statistically significantly lower than that duct the surveys. For nonclinical departments, it was an of the two comparison groups. Yet our performance on eye-opening experience. In the words of one staff member, who had never been to an inpatient unit, “Now I see this steadily improved since the inception of “Helping why my job in Finance matters!” Hippocrates” in the second quarter of 2004, and the improvement has been sustained through the most recent report (fourth quarter of 2005). Other findings were anecdotal. Staff who conducted the surveys took their responsibility seriously and enjoyed the “game” approach and the reward of the cookies. We found that, as expected, no single department bore an inordinate share of the work associated with this project. The patients who played the role of Hippocrates during the project unanimously supported it. They said that they felt glad to be part of something that would make all patients safer. Several added their own twist of humor, includFigure 4. St. Francis Hospital (SFH)’s performance on this question ing giving the Hippocrates ID band as a steadily improved since the “Helping Hippocrates” project (T-1–T-8) dursouvenir to the team that found them. ing the second quarter of 2004, and the improvement has been sustained through the fourth quarter of 2005. One “Hippocrates,” who was sitting in a
Patients’ Postdischarge Perception of Patient Identification (ID) Compliance
468
August 2006
Volume 32 Number 8
Copyright 2006 Joint Commission on Accreditation of Healthcare Organizations
Postscript St. Francis Hospital’s “Helping Hippocrates” Project was submitted to the New York State Department of Health’s 2005 Patient Safety Award. We are honored to have been the only award recipient in our category (200+ beds).9 Our $200,000 award will be used to help implement bar-coding medication administration. Our challenge now will be to ensure that all caregivers do not become complacent and
assume that bar coding has eliminated their responsibility to constantly and consistently verify patient IDs. J
Margaret A. Greenly, L.C.S.W., M.P.A., is Director, Quality Improvement and Case Management, St. Francis Hospital, Poughkeepsie, New York. Please send reprint requests to Margaret A. Greenly, [email protected].
References 1. Joint Commission on Accreditation of Healthcare Organizations: Facts About National Patient Safety Goals. http://www.jointcommission.org/ PatientSafety/NationalPatientSafetyGoals/06_npsg_facts.htm (last accessed Jun. 1, 2006). 2. Joint Commission on Accreditation of Healthcare Organizations: FAQs for the 2006 National Patient Safety Goals. http://www.joint commission.org/NR/rdonlyres/2E5C5B8C-108D-4A599303-8605B5EB3B0D/0/06_npsg_faq1.pdf (last accessed Jun. 1, 2006). 3. O’Neil K., et al.: Patient misidentification in a pediatric emergency department: Patient safety and legal perspectives. Pediatr Emerg Care 20:487–492, Jul. 2004. 4. Parisi L.: Patient identification: The foundation for a culture of patient safety. J Nurs Care Qual 18:73–79, Jan.–Mar. 2003.
August 2006
5. University of Virginia Health System: Hippocrates. http://www.health system.virginia.edu/internet/library/historical/artifacts/antiqua/ hippocrates.cfm (last accessed Jun. 1, 2006). 6. CyberNurse: Florence Nightingale Pledge. http://www.cybernurse.com/ florencepledge.html (last accessed Jun. 2, 2006). 7. Joint Commission on Accreditation of Healthcare Organizations: Speak Up Initiatives. http://www.jointcommission.org/PatientSafety/ SpeakUp/ (last accessed Jun. 1, 2006). 8. Personal Success and Leadership Institute: Joy in Work. http://pslinstitute.com/joyinwork.html (last accessed Jun. 2, 2006). 9. New York State Department of Health: Dear Administrator Letter Regarding 2006 Patient Safety Awards. http://www.health.state.ny.us/ professionals/patients/patient_safety/awards/2006_dhal_pat_safety_a ward.htm (last accessed Jun. 2, 2006).
Volume 32 Number 8
Copyright 2006 Joint Commission on Accreditation of Healthcare Organizations
469