| JOURNAL HOME | CME HOME | THIS MONTH | PAST ISSUES | ETOC | COLLECTIONS |
| AUTHORS | REVIEWERS | EDITORIAL BOARD | FEEDBACK | RSS | HELP |
| ||||||||||||||
| ||||||||||||||
|
|
|
|
||||||||||||
|
||||||||||||||
From the *Departments of Anesthesiology, Critical Care and Pain Management; and Hospital Administration, Meir Medical Center, Kfar Saba, The Sackler School of Medicine, Tel Aviv University, Tel Aviv, Israel.
Address correspondence and reprint requests to Brian Fredman, MB BCh, Department of Anesthesiology and Intensive Care, Meir Medical Center, Kfar Saba 44281, Israel. Address e-mail to Fredman.Brian{at}clalit.org.il.
Abstract
BACKGROUND: It is important to ensure a patient-safe environment in the perioperative setting. With this in mind, a "patient-safety first" philosophy was adopted within our operating room service.
METHODS: During the first phase of the interventional study (2001–2002), we defined and executed the organizational and educational aspects of the intervention. Thereafter, the implementation phase (2003–2005) was performed. According to our zero tolerance policy, in the event that a major error in patient readiness for anesthesia and surgery was found in the operating room holding area, the patient would be returned to the parent department ("failure") and the surgical procedure delayed until the major error was corrected.
RESULTS: The data of 15,856 patients were recorded. During the 3-yr implementation period, 112 patients (0.71%) were returned to the department. A statistically significant (P < 0.002) reduction in major errors was recorded when comparing the year 2003 to the years 2004 and 2005 (1.04, 0.59, and 0.49% for the years 2003, 2004, and 2005, respectively). Furthermore, stepwise logistic regression demonstrated a time-dependant significant decrease in the incidence of a major error that resulted from inadequate patient preparation (odds ratio = 1.48, 95% CI: 1.16–1.87). In addition, the mean time between failures was 6.6, 11.2, and 14.7 days for the years 2003, 2004, and 2005, respectively (P < 0.03). Finally, a significant (P < 0.0001) improvement in patient preparation over time, as well as the overall probability that the patient preparation score = 100% (P < 0.001), were demonstrated.
CONCLUSIONS: Education and increased awareness can decrease perioperative errors. However, even with a carefully designed policy in place, an error-free environment was not achieved. Therefore, monitoring and system analysis should be performed on a continuing basis.
All patient populations are at risk for administrative problems or human error (1–3). However, because of the number of different professional disciplines involved (anesthesiologists, surgeons, nursing, and ancillary staff), absolute number of required personnel, and variability of surgical procedures, the operating room (OR) environment is threatening to patient safety (4). Since ORs provide facilities to thousands of patients annually, the process of ensuring perioperative patient safety requires an organizational commitment that includes all personnel, from management to the individual health care provider. However, there have been a few interventional studies describing perioperative patient safety.
In the year 2001, a patient-safety project was initiated within the OR service of the Meir Medical Center. As a result, a "patient-safety first" philosophy was adopted and perioperative patient safety was considered to be of paramount importance in all routine clinical practice.
This report describes the methodology and results of our multifaceted, cross-organizational, interventional study that was designed to improve perioperative patient safety in a busy general hospital.
METHODS
The Meir Medical Center is a 713 bed general hospital. Fifteen independent surgical departments perform approximately 5000 "day" (08H00–16H00) surgical cases per year.
As part of our patient-safety first philosophy, the OR management initiated a patient-safety project that declared the process of preparing a patient for surgery to be the cornerstone of patient safety in the OR setting. The study was performed in two phases. During the first phase (2001–2002), the organizational and educational aspects of the intervention were defined and executed. Thereafter, the implementation phase (2003–2005) was performed.
Perioperative Checklist (Appendix I)
A designated committee that represented hospital, OR, and surgical department managements designed a preoperative readiness checklist. The checklist was constructed in such a manner as to provide a dedicated section to each of the various links in the perioperative chain of events. As the patient was transferred from station to station the professional staff completed the section describing the preparation process of that particular station. As a result, the checklist was cross-organizational and tracked patient readiness for surgery at a multidisciplinary level.
The checklist was completed by the nurse at every station. However, the doctor was responsible for performing certain functions before the patient being ready for surgery (marking the surgical side, receiving informed consent, prescribing antibiotics, etc.). It was the nurses’ responsibility to check and record that these facets of patient preparation had been completed before OR transfer.
Caregiver Education
Before initiating the checklist, all physicians (surgeons, anesthesiologists) and nurses (departmental and OR) were educated as to the importance of the checklist, its purpose and content (predetermined readiness criteria), and the technical aspects regarding its usage.
Handing Over the "Baton"
During the perioperative period, definite and specific stations were identified. These stations included the surgical department, OR holding area, and the OR. On the completion of one stage and before proceeding to the next stage of this predefined route, a dedicated member of staff completed the relevant section of the checklist and the form was stamped and signed. On arrival at the next station, the completed section of the checklist was examined and the relevant new section completed and similarly stamped and signed. This cycle of checking and rechecking was performed at each of the predetermined stages during the perioperative period.
Error Management
During the education phase of the intervention program, our zero tolerance policy regarding patient readiness for surgery was explained. This policy dictated that the predetermined checklist would be completed for every patient undergoing surgery. Furthermore, it was explained that in the event that a patient arrived in the OR holding area and a major error (Table 1) in patient readiness for anesthesia and surgery was found, the patient would be returned to the parent department and the surgical procedure delayed until the major error was corrected.
|
During the implementation phase of our study, using the predefined checklist, the adequacy of the surgical department’s patient preparation was assessed. When a major error was discovered, the patient was returned to the surgical department and the preparation process was repeated. Thereafter, the patient was returned to the OR holding area, the checklist reexamined and, if correct, the patient was transferred to the OR. On arrival in the OR, the holding area checklist was examined, and the OR section of the checklist completed. After the checking and rechecking process, surgery was performed. All events of this nature were documented. Such events were defined as "failures" and the time between failures was recorded and analyzed. Furthermore, a standing investigation committee, consisting of the hospital’s Chief Executive Officer and hospital head nurse, investigated all cases. These investigations took place within seven working days of the incident and were performed in the presence of the relevant departmental chairperson and departmental head nurse. Finally, corrections in work processes were instituted as appropriate.
Patient Preparation Score
Patient preparation score was divided into "nursing preparation" and "medical preparation." Using a categorical score (yes, no, not relevant), a dedicated nurse assessed patient preparation. Five surgical patients were randomly assessed for each department every month. Using a computer-based program, all "not relevant" data were disregarded. Thereafter, the numbers of "yes" entries were calculated as a percentage of all relevant entries.
The results of this monthly screening were made available to the OR and hospital administrations as well as to the relevant departmental chairperson and head nurse.
Statistical Analyses
Data are expressed as numbers or percentages. In all cases, normality was assessed with the Kolmogorov– Smirnov test (using the Lilliefors’ modification). Depending upon the results of the Kolmogorov–Smirnov analysis, either parametric or nonparametric analysis was performed. The number of patients returned to the surgical department was analyzed using the 
2 test. Stepwise logistic regression was performed to assess the effect of time on the occurrence of an event that resulted in a patient being returned to the surgical department due to inadequate preparation (failure). The mean time between failures was analyzed using Kruskal–Wallis test.
Patient preparation score was analyzed using 2 test for pooled data and Fisher’s Exact test for consecutive years.
The probability that the patient preparation score = 100% was assessed using one- or two-way ANOVA with Tukey method for multiple comparisons as appropriate. This assessment was preformed for all departments (pooled data) and within each department.
GraphPad Prism (version 4) or SPSS (version 14) were used to analyze the data. In all cases, P < 0.05 was considered to indicate statistical significance.
RESULTS
During the implementation phase of the study (2003–2005), patient preparation for surgery was analyzed in 15,856 patients (Table 2). A total of 112 patients (0.71%) were returned to the department during the 3-yr implementation period. A statistically significant (P < 0.002) reduction in major errors was recorded when comparing the year 2003 to the years 2004 and 2005 (1.04, 0.59, and 0.49% for the years 2003, 2004, and 2005, respectively). Stepwise logistic regression showed a significant decrease in the odds over time that a patient would be returned to the surgical department due to a major error in patient preparation (odds ratio = 1.48, 95% CI: 1.16–1.87). The mean time between failures was 6.6, 11.2, and 14.7 days for the years 2003, 2004, and 2005, respectively (P < 0.03) (Table 2).
|
Table 3 describes the incidence of major errors. The most common errors in the year 2003 were the absence or incorrect identification of side of surgery as well as the absence or usage of the incorrect informed consent form. However, the incidence of these errors significantly (P < 0.05) decreased in subsequent years. Furthermore, the incidence of an absent or incorrect identification bracelet significantly (P < 0.03) decreased between the years 2004 and 2003. By contrast, the incidence of finding incorrect documents in patient’s medical chart was significantly higher in the year 2004 compared with that in the year 2003 (P < 0.03).
|
The influence of our intervention on the patient preparation score was assessed for the surgical departments that operated on several days during the week. Consequently, the data from 10 of the 15 surgical departments were analyzed. When pooling patient preparation data for all 10 contributing departments, a significant improvement in patient preparation over time was demonstrated for consecutive years (Fig. 1).
|
The overall probability that the patient preparation score = 100% demonstrated a significant (P < 0.001) improvement over time. In addition, when this parameter was assessed within each department, a significant improvement was demonstrated in nine departments and a significant deterioration was revealed in one department (Table 4).
|
DISCUSSION
In 2003, the Joint Commission on Accreditation of Health Care Organizations published its Universal Protocol for Preventing Wrong Site, Wrong Procedure, Wrong Person Surgery. During that year, the Joint Commission on Accreditation of Health Care Organizations events trends demonstrated that approximately 15% of the total events occurred perioperatively. Furthermore, 70 wrong-site surgery events were reported. Although many causes may explain these findings, communication and orientation or training were cited as the major causes of these preoperative accidents (5).
The results of our current study demonstrate the importance of a patient-safety first philosophy. When compared with baseline data, a significant decrease in the incidence of most of the major errors was recorded. Consequently, the number of patients returned to the surgical department due to a major error was significantly decreased. However, unlike the other variables measured, the incidence of finding incorrect documents (identification labels, electrocardiogram, laboratory results, etc.) in patient’s medical chart increased as the study progressed (Table 3).
In the absence of a control group, it is important to acknowledge the possibility that our results may have been influenced by nonrelated variables (e.g., change in quality of staff, technological improvements, caregiver:patient ratios, etc.). However, despite a detailed audit designed to discover possible time-related changes in these variables, no such occurrences could be isolated. Furthermore, a significant decrease in the odds over time that a patient would be returned to the surgical department due to a major error in patient preparation, as well as a statistically significant increase in the mean time between failures, was demonstrated as the study progressed. In addition, data analysis of patient preparation score during the implementation phase of our study reveals that, when comparing consecutive years, a statistically significant improvement was noted for all years (Fig. 1). The impact of our intervention on patient preparation is further supported by the fact that two-way ANOVA with Tukey method for multiple comparisons revealed a significant change in the probability that the patient preparation score = 100% within each department over time. This change represents an improvement in patient preparation over time in nine departments and deterioration in patient preparation in one department (Department 7, Table 4). Therefore, we suggest that the major improvements in patient preparation are the direct result of our intervention and reflect our unequivocal commitment to patient safety.
The apparent deterioration noted in Department 7 may be explained by the high level of awareness of patient safety and related issues in that department before our intervention. While the probability values recorded demonstrate a statistically significant unfavorable change, this negative trend was of no clinical importance (0.9805 in 2001 vs 0.9602 in 2005) (Table 4).
Although returning a patient to the surgical department has financial implications, the hospital administration considered patient readiness for surgery to be of ultimate importance. This, together with our structured system for patient preparation for surgery, facilitated routine management of our patient-safety-related variables. However, although all members of the greater surgical team underwent training and assumed a full partnership in the endeavor to ensure patient safety, an error-free environment was not achieved. This phenomenon is likely multifactorial. First, because of cost concerns, patient-safety-related tasks were not performed by dedicated or additional staff. As a result, during routine day-to-day practice, patient safety was not always a high priority. Second, since our management system relied heavily on human effort, errors could not be totally eradicated by rules, checking, and rechecking. On the contrary, checking can become ritualistic with the emphasis placed on performing a series of mechanical and depersonalized steps (6). Third, our perioperative chain of events was characterized by each station checking similar variables to that of the station that preceded it. As a result, it is possible that aspects of patient preparation were overlooked on the assumption that the error would be corrected before the patient actually being placed on the OR table. Finally, an additional potential source of error relates to turnovers in staff. Although during the organizational and educational phase (2001–2002) of our intervention, all staff was educated as to the importance of the checklist, its purpose and content, and the technical aspects regarding its use, maintenance of this knowledge and its dissemination to new staff members was the responsibility of the departmental chairperson and head nurse. Thus, it is possible that ignorance or inexperience contributed to the fact that a totally error-free environment was not achieved. However, this hypothesis requires further investigation.
The ritual of returning a patient to the surgical department was crucial to the process and was associated with multiple beneficial effects. First, all caregivers understood that the patient-safety first policy was serious and not subject to compromise. Second, despite our nonthreatening environment, the need to explain the cause and reason for the error before an investigational committee was a deterrent for further error. Third, explaining the events to patients and their families was similarly uncomfortable. Consequently, patient safety became a central axis around which the entire surgical division revolved.
In conclusion, although all health caregivers understand the need for a patient-safety program, during routine day-to-day practice this subject is often neglected. As a result, patients may be subjected to unnecessary risk. Consequently, increased awareness of safety-related issues is essential. Furthermore, a structured program and perioperative checklist are effective administrative aids. However, since an error-free environment is likely unattainable and human error inevitable, education, monitoring and system analysis must be performed on a continuing basis.
|
Footnotes
Accepted for publication April 5, 2007.
REFERENCES
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
0.0002 when comparing consecutive years except for years 2001 and 2003 (Fisher’s exact test).
