Key points

Background: checklists started in the aeronautic industry

In the summer of 1934, the US Army Air Corps circulated a proposal for a new long-range bomber to replace the 2-engined B-10, which was currently in use. Prospective builders were instructed to have multiengined aircraft ready for a competition in October 1935. The candidate aircraft had to be able to fly at least 1640 km (1020 miles) and preferably 3540 km (2200 miles). They had to be able to carry a 900-kg (2000-pound) bomb load and to be able to reach a speed of at least 320 kph (200 mph), although 400 kph (250 mph) was considered desirable.

Working in secrecy, Boeing produced a prototype, the Model 299. When a Seattle newspaperman saw the prototype, he named it a “flying fortress”; the name stuck. The Model 299 had 4 engines, rather than 2 or 3; retractable landing gear; electric trim tabs on its control surfaces; a hydraulically operated constant-speed propeller; and positions on the fuselage for gun turrets. It was a more complicated plane than the B-10 and was the first 4-engined plane ever built.

After a short period of testing the 299 was delivered to Wright Field, Ohio, for testing against a Martin design, an upgraded B-10, and a DC-2 Douglas converted into a bomber, the DB-1. Both were good designs, but were 2-engined aircraft. Boeing’s 299 Flying Fortress was in a class by itself. It could carry 5 tons of bombs, depending on the fuel load, which was far more than its 2-engined competitors; the 299 carried its load higher, faster, and nearly twice as far as its competitors.

On October 30, 1935, the Fortress prototype taxied out for takeoff at Wright Field. A crowd gathered to watch. At the controls was the Air Corps’ chief test pilot, Major Ployer P. Hill. His copilot was First Lt. Donald L. Putt. Also aboard were an engineer, a mechanic (both were in the rear) and Leslie R. Tower, the Boeing test pilot, who was standing in the cockpit behind the two pilots.

The aircraft roared down the runway and took off. It then climbed steeply-too steeply. It rose to an altitude of about 90 m (300 ft), where it stalled, rolled to the side, crashed back onto the airfield and exploded. Putt and Tower stumbled out of the wreckage dazed and bleeding. The two mechanics went out the back, largely unscathed. Hill was unconscious and trapped in the cockpit. He was evacuated from the wreckage but died the next day. Tower, who had been standing behind the pilots as an observer, blamed himself for the accident. Although he did not seem to be seriously injured, he died not long afterward.

Investigators determined that the Fortress crashed because the elevator and rudder controls were locked; the pilot could not lower the nose, so the aircraft quickly stalled. The locking mechanism was controlled from inside the cockpit, but no one remembered to disengage it before takeoff. Tower apparently noticed that the control lock was still engaged as the aircraft moved up to stall, but was unable to get to it in time to prevent a crash. More familiar with the 299 than anyone else, this oversight on his part is why he blamed himself for the disaster. Because the Boeing prototype had crashed, the Corps declared the winner to be the Douglas DB-1, later designated the B-18 Bolo.

Air Corps leaders tried to place an order for 65 of the revolutionary Fortresses, but War Department General Staff, who controlled Air Corps finances, refused. The General Staff advanced the view that, because the Boeing airplane had crashed, it must have been too complex for anyone to handle safely. Acting on the misguided principle that quantity was more important than quality, the Army promptly ordered 133 of the new Bolos.

A group of test pilots thought that the Flying Fortress, although complex, was flyable. They came up with a checklist for pilots to use before take off, while taxiing, during flight, and landing, to ensure that some simple but crucial step, such as unlocking the elevator and rudder controls, had not been forgotten. Through a legal loophole, the Air Corps was eventually able to purchase 13 Flying Fortresses, enough to equip 1 squadron. These planes were designated YB-17s. Using the checklists, Air Corps pilots logged more than 9200 flying hours on their YB-17s without experiencing a serious major accident.

When World War II broke out in Europe in September 1939, the Army Air Corps had barely 24 of the new B-17s. In September 1940, the number was up to only 49 bombers. The United States needed to increase production, but things still moved at a glacial pace. At the time of Japan’s attack on Pearl Harbor on December 7, 1941, the Air Corps had fewer than 200 B-17s in the inventory. Not until early 1944 would the US military have enough Fortresses to have a decisive impact on the bombing campaign against Germany. The Army eventually purchased about 13,000 Flying Fortresses. Three-hundred and fifty Bolos were purchased. They proved unsatisfactory in combat and were relegated to coastal patrols and navigational training.

The 1935 crash did produce one notable benefit. Airmen realized that aircraft were becoming too complex to fly safely without standardized procedures. Moreover, these procedures were too numerous and complicated to commit entirely to memory. Checklists were developed that spelled out specific tasks that were to be accomplished by each crew member at various times throughout the flight and also while on the ground. Such a checklist, performed while taxiing out for takeoff, would probably have revealed that the 299’s elevator locks were still engaged. Today, such detailed checklists are mandatory for all aircraft.

Background: checklists started in the aeronautic industry

In the summer of 1934, the US Army Air Corps circulated a proposal for a new long-range bomber to replace the 2-engined B-10, which was currently in use. Prospective builders were instructed to have multiengined aircraft ready for a competition in October 1935. The candidate aircraft had to be able to fly at least 1640 km (1020 miles) and preferably 3540 km (2200 miles). They had to be able to carry a 900-kg (2000-pound) bomb load and to be able to reach a speed of at least 320 kph (200 mph), although 400 kph (250 mph) was considered desirable.

Working in secrecy, Boeing produced a prototype, the Model 299. When a Seattle newspaperman saw the prototype, he named it a “flying fortress”; the name stuck. The Model 299 had 4 engines, rather than 2 or 3; retractable landing gear; electric trim tabs on its control surfaces; a hydraulically operated constant-speed propeller; and positions on the fuselage for gun turrets. It was a more complicated plane than the B-10 and was the first 4-engined plane ever built.

After a short period of testing the 299 was delivered to Wright Field, Ohio, for testing against a Martin design, an upgraded B-10, and a DC-2 Douglas converted into a bomber, the DB-1. Both were good designs, but were 2-engined aircraft. Boeing’s 299 Flying Fortress was in a class by itself. It could carry 5 tons of bombs, depending on the fuel load, which was far more than its 2-engined competitors; the 299 carried its load higher, faster, and nearly twice as far as its competitors.

On October 30, 1935, the Fortress prototype taxied out for takeoff at Wright Field. A crowd gathered to watch. At the controls was the Air Corps’ chief test pilot, Major Ployer P. Hill. His copilot was First Lt. Donald L. Putt. Also aboard were an engineer, a mechanic (both were in the rear) and Leslie R. Tower, the Boeing test pilot, who was standing in the cockpit behind the two pilots.

The aircraft roared down the runway and took off. It then climbed steeply-too steeply. It rose to an altitude of about 90 m (300 ft), where it stalled, rolled to the side, crashed back onto the airfield and exploded. Putt and Tower stumbled out of the wreckage dazed and bleeding. The two mechanics went out the back, largely unscathed. Hill was unconscious and trapped in the cockpit. He was evacuated from the wreckage but died the next day. Tower, who had been standing behind the pilots as an observer, blamed himself for the accident. Although he did not seem to be seriously injured, he died not long afterward.

Investigators determined that the Fortress crashed because the elevator and rudder controls were locked; the pilot could not lower the nose, so the aircraft quickly stalled. The locking mechanism was controlled from inside the cockpit, but no one remembered to disengage it before takeoff. Tower apparently noticed that the control lock was still engaged as the aircraft moved up to stall, but was unable to get to it in time to prevent a crash. More familiar with the 299 than anyone else, this oversight on his part is why he blamed himself for the disaster. Because the Boeing prototype had crashed, the Corps declared the winner to be the Douglas DB-1, later designated the B-18 Bolo.

Air Corps leaders tried to place an order for 65 of the revolutionary Fortresses, but War Department General Staff, who controlled Air Corps finances, refused. The General Staff advanced the view that, because the Boeing airplane had crashed, it must have been too complex for anyone to handle safely. Acting on the misguided principle that quantity was more important than quality, the Army promptly ordered 133 of the new Bolos.

A group of test pilots thought that the Flying Fortress, although complex, was flyable. They came up with a checklist for pilots to use before take off, while taxiing, during flight, and landing, to ensure that some simple but crucial step, such as unlocking the elevator and rudder controls, had not been forgotten. Through a legal loophole, the Air Corps was eventually able to purchase 13 Flying Fortresses, enough to equip 1 squadron. These planes were designated YB-17s. Using the checklists, Air Corps pilots logged more than 9200 flying hours on their YB-17s without experiencing a serious major accident.

When World War II broke out in Europe in September 1939, the Army Air Corps had barely 24 of the new B-17s. In September 1940, the number was up to only 49 bombers. The United States needed to increase production, but things still moved at a glacial pace. At the time of Japan’s attack on Pearl Harbor on December 7, 1941, the Air Corps had fewer than 200 B-17s in the inventory. Not until early 1944 would the US military have enough Fortresses to have a decisive impact on the bombing campaign against Germany. The Army eventually purchased about 13,000 Flying Fortresses. Three-hundred and fifty Bolos were purchased. They proved unsatisfactory in combat and were relegated to coastal patrols and navigational training.

The 1935 crash did produce one notable benefit. Airmen realized that aircraft were becoming too complex to fly safely without standardized procedures. Moreover, these procedures were too numerous and complicated to commit entirely to memory. Checklists were developed that spelled out specific tasks that were to be accomplished by each crew member at various times throughout the flight and also while on the ground. Such a checklist, performed while taxiing out for takeoff, would probably have revealed that the 299’s elevator locks were still engaged. Today, such detailed checklists are mandatory for all aircraft.

Lessons of the aeronautic industry extended to health care: a checklist and bloodstream infections

It may seem strange to try to adapt techniques devised to make flying complicated aircraft safe to the practice of medicine, but, in 2001, a physician at Johns Hopkins Hospital, Peter Pronovost, PhD, MD, decided to try to put one together to decrease the rate of complications of one of the tasks that physicians do daily in most hospitals: placement of central lines.

In 2006 it was reported that 36 million patients were admitted to hospitals in the United States, staying for 164 million hospital days. Eleven percent of those hospital days are spent in intensive care units (ICUs), or 9.7 million days; for 54% of the days (9.7 million), central venous catheters remain in place to infuse medicine and fluids. At that time there were 48,600 catheter-related bloodstream infections resulting in deaths estimated from 17,000 to 28,000 per year. The median rate of catheter-related bloodstream infections in ICUs ranged from 1.8 to 5.2 per 1000 catheter days.

The intervention used evidence-based procedures recommended by the Communicable Disease Center as having the greatest effect on decreasing the rate of catheter-related bloodstream infection. These procedures were that physicians wash their hands before the catheter placement; full barrier protection is placed on the patient before insertion of the catheter; the physician wears sterile gloves, mask, hat and gown; the skin of the patient is scrubbed with chlorhexidine; the femoral site should be avoided, if possible; and unnecessary catheters should be removed as soon as possible. Dr Pronovost devised a 1-page checklist to ensure that these tasks were performed. Nurses stopped providers in nonemergency situations from proceeding with catheter placement if the steps were not followed.

This checklist was tried out at Johns Hopkins Hospital; results were dramatic: the 10-day infection rate went from 11% to 0%. Pronovost then devised checklists to ensure that nurses observed patients for pain at least once every 4 hours, which reduced the likelihood of patients enduring pain from 41% to 3%. Another checklist ensured that patients on mechanical ventilators received antacid medication and that the head of the bed was propped up to at least 30°.

The percentage of patients not receiving antacids went from 70% to 4%; the incidence of pneumonia decreased about 25%. Checklists helped with memory recall, established the minimum necessary steps in a process, and established a higher standard of baseline performance.

The checklist to reduce catheter-induced infections was introduced in most of the ICUs in Michigan as part of a statewide safety initiative known as the Michigan Health and Hospital Association (MHA) Patient Safety and Quality Keystone Center ICU project. The project also introduced a daily goals sheet to improve clinician-to-clinician communication within the ICU, an intervention to reduce ventilator-assisted pneumonia, and a comprehensive unit-based safety program to improve safety culture. The project involved 67 hospitals, of which 52% were teaching facilities and included 85% of all of the ICU beds in Michigan.

Data were collected from 103 ICUs for 1981 ICU months and 375,757 catheter days. Using the checklists, the overall median rate of catheter-related bloodstream infection decreased from 2.7 (mean 7.7) infections per 1000 catheter days at baseline to 0 (mean 2.3) at 0 to 3 months after implementation of the study intervention, and was sustained at 0 (mean 1.4) during 18 months of follow-up. Teaching and nonteaching hospitals realized similar improvements.

These data were published in the New England Journal of Medicine . An editorial in the same issue discussing this article stated that “the story is compelling and the costs and efforts so relatively minor that the five components of the intervention should be widely adopted. We can no longer accept the variations in safety culture, behavior or systems of practice that have plagued medical care for decades. Imagine the effect if all 6000 acute care hospitals in the United States were to show a similar commitment and discipline.”

Development of the World Health Organization checklist

In an article published in 2008, Weiser and colleagues, reported that the World Health Organization (WHO) had collected demographic, economic, and health data from the 192 WHO member states. WHO estimated that 232 million major surgical procedures are performed each year. The article concluded that “Worldwide volume of surgery is large. In view of the high death and complication rates of major surgical procedures, surgical safety should now be a substantial global public-health concern. The disproportionate scarcity of surgical access in low-income settings suggests a large unaddressed disease burden worldwide. Public-health efforts and surveillance in surgery should be established.”

In January 2007 in Geneva, Switzerland, the first meeting of Safe Surgery Saves Lives convened for a 2-day conference, bringing together surgeons, anesthesiologists, nurses, hospital administrators, and others to improve the safety of surgery worldwide and to obtain better information on the nature of surgical services in different countries and in different health systems.

The group concluded that a surgical checklist should be developed. The checklist should ensure that proper antibiotics were given before incising the skin and that monitored anesthesia was administered. The checklist would emphasize teamwork and be occupied with measures that promote safety. It should include a preoperative briefing to address surgical team issues and also be a team training process. The checklist should facilitate teamwork. Members at the conference in Geneva recognized that different countries and different specialties would have different needs; the checklist should therefore provide latitude for additions and tailoring based on local factors and environment. The checklist that was developed as a product of this conference and working sessions that followed is available at www.safesurgery.org and www.who.int/patientsafety/safesurgery/tools .

The WHO checklist contained 19 items to be noted before and after surgery: that patients confirmed their identity, surgical site, and procedure, and that a consent was signed; if applicable, the surgical site was marked; a pulse oximeter was present and functioning; members of the team were aware if the patient had a drug allergy; airway had been evaluated; and, if blood loss of at least 500 mL was expected, blood and fluids were available. The goal was to create a tool that supported clinical practice without substituting a rigid algorithm for professional judgment. Following the aviation lesson, the checklist was to focus on items that are recognized to either be deadly if missed or, if not deadly, then high risk and known to be recurrently overlooked or missed.

In the WHO checklist, a time-out is performed before skin incision. The patient’s name, surgical site, and procedure are reviewed. All team members are identified by name and role; surgical, anesthesia, and nursing staff review the anticipated events and confirm that preoperative antibiotics have been administered. All imaging studies for the correct patient are displayed in the operating room, if necessary. Following surgery, the nurse reviews the name of the procedure and that needle, sponge, and instrument counts were correct. Any specimen, if necessary, has been labeled. Issues with equipment are addressed.

Between October 2007 and September 2008 8 hospitals in 8 cities (Toronto, Canada; New Delhi, India; Amman, Jordan; Auckland, New Zealand; Manila, Philippines; Ifakara, Tanzania; London, United Kingdom; and Seattle, WA) participated in the WHO’s Safe Surgery Saves Lives program. Selection of these cities purposely included places with different economic circumstances and different populations. The checklist was introduced into these hospitals, each of which had a full-time investigator for the project with no other clinical responsibilities. Each hospital identified 1 to 4 operating rooms to serve as study rooms. Patients who were 16 years of age or older and were undergoing noncardiac surgery were consecutively enrolled in the study. After noting the practices at that time in each institution, all were asked to correct policies not consistent with the 19-item WHO safe-surgery checklist and to implement the checklist in the designated rooms. Part of the data was collected by observers in the operating room and part by clinical teams involved in surgical care.

During the baseline period 3733 patients were enrolled; 3955 patients were enrolled after the checklist was implemented. The rate of complications decreased from 11% at baseline to 7% after the checklist was introduced. The total in-hospital rate of death decreased from 1.5% to 0.8%. These decreases were of about 36%. Similar declines in complications were observed in high-income and in low-income sites. It was noted that, “The rates of reduction in rates of death and complications suggest that the checklist program can improve the safety of surgical patients in diverse clinical and economic environments.”

There have been some legitimate questions raised about the findings of the 8-hospital WHO study. Martin and colleagues thought that a 30% reduction in death was unlikely to be achieved in the United Kingdom because rates of death in some hospitals in the WHO exceeded the published normal range of 0.4% to 0.8%. McCambridge and colleagues noted that clinical teams were aware that they were being observed and that some of the improved outcomes may have been influenced by alterations in behavior. Sanders and Jameson thought it was possible that antibiotics and pulse oximetry may have accounted for the survival advantage in the sites in cities of low income. In response to these doubts, Haynes and Gwande pointed out that the case mix varied widely among hospitals and that the hospitals had enormous diversity. Rate of postoperative death is unknown for the mix of cases in this international group of hospitals and comparison of these hospitals with those in developed countries is invalid. WHO recommends that the use of an oximeter and antibiotics are minimum standards for safe surgery. Haynes and Gwande found no effect of an observer in the operating rooms.

Many of the findings of the WHO Safe Surgery Saves Lives study were confirmed in a tertiary university hospital in Utrecht, the Netherlands.