Last month’s annual meeting of the Association of Air Medical Services was its usual low-key success as some 2,500 aeromedical professionals, a record number, assembled in the Kansas City (Mo.) Convention Center for three days of conferences and trade show exhibition. In a gathering of what is probably the most highly networked aviation environment outside of the military, old friends met, new friends were made and deals were struck within the framework of an ambitious schedule of seminars, lectures, professional development courses and the presentation of scholarly papers.
But what was on everyone’s mind was safety, thoughts centering on the alarming recent increase in aeromedical accidents, an echo of previous upswings in the accident rate in decades past. In 1998 the industry saw the beginning of a significant upswing in the numbers of aeromed accidents. Beginning with nine mishaps in 1998 and trending upward to 13 accidents last year (data for 2002, while partial, shows five accidents recorded as of September 30), those accidents have resulted in 52 fatalities. The 13 fatalities in both 1998 and this year so far make for the deadliest years for aeromed operations since 1991, when 17 fatalities were recorded.
Many theories as to the recent accident upswing were bandied about among the rank and file at AAMS, echoing the traditional safety areas of concern in the past: the need for pilots to pick up patients from unfamiliar, often off-road landing sites; the need to fly in inclement weather, (because that’s when the accidents that create accident victims in need of help often happen); and the perception that helicopters suffer mechanical failures more often than their fixed-wing brethren.
The Truth Is Out There
To find out just what is happening with aeromed safety, a team of researchers from the University of Chicago Aeromedical Network (UCAN) safety committee in fall 2000 began an independent investigation of the situation. Last month in Kansas City the group presented its insightful but sometimes surprising findings.
The UCAN team was headed by Dr. Ira Blumen, who describes his group’s search for the truth thus: “For more than a decade, when our industry and its observers looked at EMS accident information, the only available information was the number of accidents and fatalities–commonly referred to as the raw data. No one was able to determine if the increase in accidents was simply related to an increase in the number of hours flown or whether helicopter EMS [HEMS] had indeed become more dangerous.
“Our primary investigation began with an extensive review of accident and incident data specific to the HEMS industry. Based on this review, UCAN developed an exposure data model consisting mainly of the number of total transport flights and the length of the transports. Without this exposure data it would be impossible to calculate HEMS accident and fatality rates or draw any meaningful conclusions or comparisons.
“The study focus next shifted to a comparison of HEMS accident and fatality rates to other forms of air travel, followed by a comparison of air medical transport with other occupations and their ‘routine’ risks to contrast the fatality rates and odds of death. To accomplish this, the population at risk in HEMS would need to be determined if we were to attempt to make these unique comparisons. This data has never been tracked or even estimated in the HEMS literature.”
As presented by the UCAN researchers:
• Since 1972, HEMS operations have flown an estimated three million hours while transporting 2.75 million patients. In the 31 years preceding last September, there have been 162 accidents involving dedicated medical helicopters and four involving dual-purpose helicopters in the U.S. Of those 162 accidents, 67 involved fatalities, claiming the lives of 144 crewmembers.
• Since 1998 there have been 50 accidents– nearly 31 percent of all the accidents recorded over three decades. However, the percentage of fatal accidents has declined, down nearly 30 percent compared with the early 1980s. Nevertheless, since 1990 there have been on average 2.5 fatal accidents annually, claiming the lives of five or six crewmembers a year.
• Night ops are more dangerous. Only 38 percent of all HEMS flights are at night, but 49 percent of the accidents recorded between 1978 and 1998 took place at night.
• Takeoffs and landings have traditionally been assumed to be the most hazardous parts of any flight. Not so, according to the UCAN data, which showed that more accidents (36 percent) occurred during cruise flight than any other mission phase.
• Human factors were associated with 65 to 76 percent of the accidents studied. In fatal accidents, human error was cited as a cause in 84 percent of the accidents recorded. Human error was cited as a factor in more than two-thirds of the en route accidents, more than 80 percent of the accidents during takeoff and approximately 90 percent of the accidents recorded during approach and landing. (It should be stressed that this ratio of human error adheres to a general 20/80-percent ratio that holds true through most non-aviation accidents, 20 percent of which are caused by machine-related failures, with the remaining 80 percent traceable to human error. “Even assuming this 20/80 split is a way of life throughout most of human endeavor,” mused Blumen, “that leaves no logical reason for the upswing in lethal HEMS accidents in recent years.”)
UCAN researchers combed the available data for clues that could lead them to detect trends leading to useful conclusions. Some facts were undeniable, even though they were bewildering. By a wide margin, maintenance-related accidents took place more during cruise flight, rather than during landing and takeoff, when mechanical stresses are at their peak.
Pilot experience data offered up an interesting paradox. Pilots with anywhere from 3,001 to 6,000 hr logged were by far the likeliest to have an accident. Were it not for another fact the UCAN searchers dug out of the data, one might well wonder about the advantages of experience: in 27 of the 122 accidents (22 percent) examined in a 1999 study that presented the best compilation of flight hour/accident data seen to date, the pilot-in-command had fewer than 200 flight hours in the make and model of aircraft. Eighteen (15 percent) had fewer than 100 hr and one had only three hours. So there would seem to be a solid link between time in type and safety.
The debate between the relative safety of single- vs twin-engine aircraft has raged for years. Of the 17 accidents attributed to engine problems, 12 involved singles and three befell twins. However, UCAN searchers concluded such comparison was meaningless because singles dominated the fleet in the early days of aeromedical transport. Subsequent data gathered by HAI indicates what should come as little surprise: twins have a lower accident rate overall than singles, whether in aeromed use or not.
With human error by far the leading cause of accidents, UCAN researchers cited a 2001 study for the Flight Safety Foundation in which author Patrick Veillette, Ph.D., examined 87 HEMS mishaps, 66 of which were attributed to human error. In this accident group, faulty in-flight decision making, inadequate preflight planning and risk taking led the other causes (failure to follow SOPs, delayed remedial actions and misinterpretation of environmental cues) by a factor of two to one.
When compared with other modes of aviation, HEMS has had its ups and downs. Twenty years ago HEMS was deadlier, by far, than any other form of non-military aviation. Then in 1990 there were no HEMS accidents, casting the mold for almost a decade of declining fatal accident rate. Since 1998, however, the fatality rate has significantly risen.
Under Pressure
NASA’s Aviation Safety Reporting System (ASRS) has for decades been a well regarded source of grassroots-level aviation safety data, and the UCAN researchers availed themselves of it, gleaning from its thousands of entries a picture of what HEMS pilots live with–and the close calls they learn from. Topping the list as a source of human factors stresses experienced and reported were:
• Communications–Coordinating the flow of information among ATC, the hospital, the health-care staff working with the patient, and ground crew, all the while flying the aircraft, has proved stressful for pilots, especially those working in the single-pilot environment. Seventy-eight percent of the aeromed pilots who submitted ASRS reports cited communications problems as the leading cause of the incidents they were reporting.
• Time pressure–Of course, getting patients where they need to be and saving time and lives is what the HEMS industry is all about. Flight preparation time is at a premium in the HEMS business and that led pilots’ time-related concerns, followed by patient condition, flight to the pickup location and low fuel. Following the first big upswing of EMS accidents in the 1980s, the practice of informing pilots of the age, sex and condition of the patient was discontinued, lest such considerations unduly influence the go/no-go decision.
• Distractions–Sophisticated in-flight aircraft equipment, the need to monitor multiple radio frequencies, traffic avoidance in high-density areas, task interruptions, poor visibility, marginal weather, noisy onboard medical equipment and an impending low-fuel situation, were all reported as contributing factors. Also listed were internal issues affecting performance, such as personal and family issues, boredom and general complacency that all too often accompanies cruise flight.
In Conclusion
“There is no typical HEMS accident,” the UCAN researchers concluded. “However, a disproportionate number of HEMS accidents occurred during night operations, during the cruise phase of flight and in the course of on-scene transports. Pilot error was flagged as the direct or indirect cause of HEMS accidents approximately three times more often than mechanical failure. Of the pilot errors, one-third were weather related.
“In 1988 the NTSB concluded that poor weather posed the greatest single threat to HEMS operations. More than a decade later, deteriorating weather conditions continue to represent a significant risk to HEMS. The cause of the weather-related accidents does not appear to be a pilot’s disregard for established weather minimums at takeoff. Instead, it is the pilot’s encounter with IMC en route. In general, weather may not cause the accident but it may increase the likelihood that an accident will happen.
“Weather is the second most common factor or cause of HEMS accidents. Of the weather-related HEMS accidents, more than 85 percent occurred at night and approximately 75 percent of those accidents resulted in fatalities. The correlation between weather-related accidents and cruise flight is very strong.
“Accordingly, an IFR rating and currency may be very helpful, if not invaluable, to overcome a situation and avoid an accident.
“In the early and mid-1980s, the accident rate for HEMS was dramatically higher than all other aviation operations. Since 1987, however, we have seen a significant decrease in the HEMS accident rate to about one-third of what we had previously experienced. The HEMS accident rate has remained consistently below the accident rate for both general aviation and all helicopter operations since the late 1980s. The fatality rate has also seen significant improvement since the late 1980s. Despite a recent increase, the fatal accident rate has been reduced by approximately 75 percent compared with the early 1980s.”
Reprints of the new publication, A Safety Review and Risk Assessment in Air Medical Transport, may be obtained from the Air Medical Physician Association, 383 F Street, Salt Lake City, UT 84103; telephone (801) 534-0829; fax (801) 534-0434; www.ampa.org.