The 20 hospitals collected 1,732 failures (median=62 failures per hospital) across 173 work areas. Hospitals conducted intervention activities in a median of five work areas. The most commonly included work areas were the ED, ICU, OR, and a Medical/Surgical unit. In 14 of 20 hospitals, senior leaders expressed that the intervention enabled them to gain a deeper understanding of the operational failures their employees encountered and the actions needed to address them effectively. One CEO stated, “I think this has been very educational for each of us, not just in what we observed but also in thinking about the decision process we used to make selections.”
There was variance in implementation effort across the hospitals. One hospital visited only three work areas, while four hospitals visited more than 10. Similarly, with regard to operational failures, the bottom quartile identified 42 or fewer failures while the top quartile recorded between 163 and 229 failures. Leadership involvement also differed. Five of the 20 hospitals successfully enlisted four or more senior managers to participate in work system observations and safety communication forums. At another nine hospitals, three senior managers engaged in the intervention activities. Four hospitals included two senior managers. The remaining two hospitals involved just one or none of the senior managers, instead giving primary responsibility to the patient safety officer and vice-presidents.
The two most frequently identified categories of operational failures were equipment/supply and facility design failures (18 percent each), which were reported in all 20 hospitals. The next three most frequently occurring failure types were communication/documentation (16 percent), staffing/staff development (16 percent), and medication (12 percent). These failures also were mentioned in nearly all participating hospitals. The remaining five categories included operational failures related to policies (5 percent), response time (4 percent), security (4 percent), infection control (3 percent), and task management (2 percent). Although these five categories occurred less frequently, they were prevalent and reported by 80–85 percent of the hospitals. summarizes the results.
Summary of Operational Failures by Category
Over half of the more than 300 equipment/supply failures were due to broken or missing equipment or supplies. This problem was pervasive: 85 percent of hospitals reported missing equipment or supplies. In the surgical unit of Hospital 122, a chronic deficiency of oxygen tanks meant patients were transferred between units without them. As a result, patients were more likely to become unstable, resulting in admission to the ICU and longer hospital stays. Our interviews with front-line staff reinforced the impact of equipment problems on patient care. For example, during our interview, a patient care technician on a busy 40-bed telemetry unit at Hospital 32 informed us of wasted time due to a chronic shortage of pulse oximeters, in part because equipment was removed for maintenance without replacement.
Another third of the equipment/supplies failures centered on equipment that did not meet the needs of the patients. A common concern was chairs and operating tables that did not support bariatric patients. Such failures represent systems problems that hinder the timely delivery of safe care.
Nearly half of the facilities failures were due to poor facility layout or lack of space. All but one hospital identified layout or space issues. Poor layout of the unit often made it difficult for providers to observe patients. Personnel also frequently complained of insufficient storage space, which often resulted in storage of dirty equipment in nondesignated places, such as hallways, creating a fire and infection control risk.
Another 22 percent of facility failures stemmed from a lack of functionality of the existing facility such as insufficient heating/cooling and facility cleanliness. Poor lighting and a lack of automatic doors hindered physicians and staff. For example, in Hospital 32's catheter lab, the overhead lighting was inadequate for procedures on patients’ left shoulders, forcing physicians to forgo the use of protective shielding because the arm it was attached to interfered with the only movable light fixture.
One-third of communication/documentation failures were the result of poor coordination among health care providers. Physicians reported inconsistent notification from nurses about changes in patient conditions, such as abnormal test results and drops in blood pressure. Nurses commented that physicians were often difficult to reach for consultation. In addition, OR staff reported that they were frequently uninformed about scheduling changes, which resulted in confusion and delay. Staff from all departments lamented a lack of advance notice about patient conditions—such as a need for supplemental oxygen or isolation—which created safety risks because they were unprepared when the patient arrived. Documentation was often redundant or unwieldy.
Communication/documentation failures also cause waste and rework. For example, Hospital 32's documentation of surgeons’ preoperative preferences was not always correct or sufficiently precise. The inaccurate system wasted money and time because staff opened unnecessary supplies and equipment.
Staffing/Staff Development Failures
Half of the staffing/staff development failures stemmed from insufficient staffing levels. Nurses reported that timeliness and quality of care suffered when the unit was understaffed. In addition, staffing shortages in ancillary departments, such as laboratory, negatively impacted patient care and safety throughout the hospital. Finally, senior managers observed that a lack of support staff (housekeepers, secretaries) caused inefficiencies and interruptions for physicians and nurses who had to perform support staff functions (e.g., answering phones, cleaning rooms) themselves. Another 30 percent of staffing failures were related to staff development. Employees requested more orientation, continuing education, in-services, and periodic review of procedures.
The most frequent medication failures centered on communication/documentation and equipment/supplies. Together, these comprised over one-half of the medication-related operational failures. Communication issues between the units and pharmacy were mentioned, such as not knowing whether the pharmacy received a fax order, resulting in extra steps to confirm receipt.
Failures related to medication equipment frequently involved automated drug dispensing (ADD) machines. These machines ran out of medications in five hospitals. Similar to studies of computerized physician order entry systems (Koppel et al. 2005
), we found that the ADD machines had discrepancies that could lead to medication errors such as a mismatch between the dose in the drawer and the dose on the screen, active patients being deleted from the computer, or patients coming up on the screen who had already been discharged or who were never on that unit. The increased complexity associated with stocking similar drugs and multiple doses of the same drug also impacted efficiency. One hospital's ICU staff complained that weekly inventory of its ADD machine took 45 minutes. The staff suggested that utilizing existing information about medication usage could facilitate substantial reductions in the number of medications stocked and hence a reduction in the time required to inventory the medication in the machines.
Policy, Response Time, Infection Control, Security, and Pace of Work
The remaining five categories cumulatively accounted for 18 percent of the operational failure data. Staff reported wait times of up to 1.5 hours for necessary supplies, such as blood products, dietary trays, and sterile equipment. There were also delays in procedures (e.g., line insertion), results from tests, and even surgeon arrival, which caused delays in treatment for patients and sometimes rework as patients waited so long they needed to be re-assessed.
Personnel also raised concerns about insufficient infection control. Poor hand washing practices were discussed at eight hospitals. Poor room layout and equipment design made it difficult to comply with infection control policies. For example, two hospitals’ senior managers observed staff placing medical equipment/supplies on a nonsterile surface, such as a patient's bed, because a clean surface was unavailable. Staff also felt vulnerable because of a perceived lack of security in the hospital, especially on the night shifts and in remote areas of the hospitals. Regarding work pace, staff most commonly complained about frequent interruptions and overlapping priorities that required them to move quickly from caring for one patient to another.