Accident investigation, or AI, is a little like politics and religion: it isn’t discussed in polite company. Ask veteran patrollers and risk managers about AI, and you’ll probably get a shy smile, a faint mumble and suddenly find yourself discussing the Red Sox bullpen.

Which isn’t surprising. Although ski areas aren’t to blame in the overwhelming majority of premise accidents, the threat of litigation—and its subsequent costs—means that ski areas must be ready to defend themselves should a skier or snowboarder decide that their boo-boo was someone else’s fault.

That’s not to say that AI reports should be created specifically to put resorts in the best possible light. The goal of an AI team is to collect data, period. Good AI teams don’t express opinions. Although team members may end up with a pretty clear idea of how an accident occurred, and may end up testifying, definitive conclusions—should a case end up in court—are generally provided by independent accident reconstruction specialists who rely on the data collected by the team.

“We look for an accurate compilation of the facts so we can reconstruct the circumstances surrounding an accident,” says Colorado attorney Peter Rietz, who defends many of the Rocky Mountains’ larger resorts and who serves as counsel to the National Ski Areas Association.

Depending on the situation, that compilation can include a wide range of components. A base area slip-and-fall is different from an on-slope collision with a hydrant, and requires different documentation. Even so, most AI reports include photographs, diagrams that can be used to accurately reconstruct the scene, information related to the path and/or actions of the injured guest prior to the accident, site measurements (along with triangulated measurements to fixed landmarks), witness statements and a variety of other elements, up to and including maintenance logs.


Building a Case

How to know what to include? Although some cases, such as a catastrophic injury, obviously warrant an immediate full-scale investigation, others aren’t so obvious. What’s more, there’s invariably a time lag between an incident and a case that reaches trial. And, as Tom Aicher of
Vermont-based Cleary, Shahi and Aicher points out, “you never know what information will ultimately prove to be important at a trial.”

Start by asking your defense attorney what he or she likes to have available. Different attorneys have different ideas about what a routine AI file should contain. “There’s a spectrum between minimalist and overdoing it,” says defense attorney Evan Hansen of Portland, Maine-based Preti Flaherty. “Realistically, you need to be able to identify where something happened and describe it accurately enough to reproduce it years down the road.”

That is no small task. In addition to the specifics of a given incident, resorts should ensure that they can find other potentially relevant information later, if needed—and this information can cover a wide range of business records. For example, parking lot slip-and-falls may be easier to defend if you have logs showing when the lot was last plowed and sanded. Terrain park incidents may be easier to defend if accurate information related to the building and maintenance of features is available. You don’t necessarily need to include these documents in an AI file, but you’d better know where they’re stored and how to retrieve them.


Act Fast

A good place to start assembling the AI file is the basic accident report form, which is realistically the first step in any AI procedure and which, if used properly, ultimately drives subsequent investigation.

“Go through the bullets of AI,” Aicher says. “The accident report is the first step. Ask yourself if a particular case warrants anything further, such as witness statements and comments. From there, decide whether you need photos and diagrams and measurements. If you complete the forms accurately and follow the proper procedures, you should cover the main issues that are likely to come up at trial.”

Few items in an accident investigation are more important than witness statements, and it’s here that resorts make the most errors. “Get as many as you can,” Rietz advises. “Ideally, you want statements from people who don’t have a horse in the race—not members of the injured skier’s party, and preferably not employees.”

Sometimes, though, that’s all you can get—and it’s much better to get those than to get nothing. Get them as quickly as possible after an incident, and don’t take them for granted. Party members’ recollections can change out of sympathy and, Hansen warns, employee statements often languish until later—if they’re collected at all.

Encourage witnesses to focus on stating the facts as they observed them. Employee witness statements are especially problematic when less-informed, line-level staffers start making judgments. “I’ve seen things like ‘the person was skiing too fast,’ or ‘was out of control,’” Rietz says. “You can’t measure speed without a radar gun, and control is subjective. It’s okay to say, ‘the person appeared to be skiing twice as fast as everyone else,’ or ‘the person appeared to be off balance.’”

It’s one thing when a veteran instructor makes a statement related to a skier’s ability or technique. But a similar statement from a lift operator or a snowmaker, even if they’re excellent skiers themselves, is easy to cast in doubt.

Hansen notes that smaller resorts tend to be more casual in collecting witness statements. “A patroller might tell a witness to ‘write down what you remember,’ rather than getting a full and formalized statement,” he says. “It’s a good idea to have a supervisor involved so relevant information is recorded accurately.”

But larger resorts aren’t immune to fuzzy reports. “Where you see sloppy stuff is when things are busy, such as a holiday weekend, when everyone is stretched,” says Hansen. “A lot of accident investigation is paperwork, and obviously first aid and mountain operation are the first priorities. On busy days, the paperwork tends to get pushed back—sometimes, too far back.”

This is why collecting data as soon as possible after an accident is essential. A ski resort is an ever-changing environment, and the most useful AI files are those that clearly define that environment at the time of the incident. That’s where photographs and diagrams come in.


A Picture’s Worth . . . $15 Million?

Here’s one bit of good news: the debate over the use of digital cameras for accident investigation is pretty much over. At one time, it was thought that use of digital images opened resorts up to accusations of photo manipulation. But digital photography in trial situations is now widely accepted. Just remember to download your images to a server as soon as you get in off the hill. Cameras disappear easily.

“Take photographs as spontaneously as possible,” says Rietz. “Emphasize the perspective that the injured party would have had.”

Diagrams are also important—and can also present problems. Defense attorneys have found that many scene diagrams contain so much information as to be nearly unintelligible. It may be better to make two or three different diagrams, rather than trying to cram everything onto one illustration. For example, one diagram might simply establish the terrain dimensions in the vicinity of an accident scene, with other illustrations showing the path skied by the injured party, the location of witnesses, and so forth. Common reference points, of course, are key. And make sure you take measurements with a tape measure or range finder. Defense attorneys still see AI files without distance measurements, and these can be essential for a successful reconstruction.

It’s not just raw data collection that’s important. The information must be reviewed promptly by supervisory personnel, to make sure that all identified witnesses have provided statements, that all relevant photos and measurements have been obtained, etc. AI files should not wait until the end of a holiday week for review; they should be assessed as quickly as possible, and if additional data is required, it should be obtained immediately.


Presenting the Case

Ski areas began doing AIs shortly after the first successful lawsuits were brought against ski resorts, and the basic tools of the trade really haven’t changed much since then (see page 68). True, laser range finders, which can be purchased for about $200, have distinct advantages over surveyor’s tape measures, and GPS receivers can be handy. But realistically, the tools and techniques involved in AI today are simply refinements over those that have been in place for decades.

There are, however, some significant advances in what can be done with the data once it’s been collected, and Brian Brill, a former Keystone patroller who now works as an accident reconstruction expert, is at the leading edge. His tools of choice start with good ol’ Microsoft PowerPoint.

Brill, who joined Keystone’s AI team in 1992 and became its top dog in 2000, was looking for better, simpler ways to collect and display the data his team collected. “I experimented with Photoshop, CorelDraw and several other applications,” he says. “I settled on PowerPoint, because although people think of it as presentation software, you can draw in it, and it was already resident on the computers we were using.”

At first, Brill simply drew maps and added in symbols: green circles for lift towers, blue diamonds for snowmaking hydrants, and so forth. The resulting diagrams were visually cleaner than the enlarged-sketches-on-foamcore that were previously used at trial. PowerPoint presentations also allowed AI photography to be displayed in a more professional manner.

But Brill didn’t stop there. He discovered that Keystone had a highly accurate vector map of the resort on its servers. Vector maps, unlike those stored in a raster format, are almost infinitely zoomable. This enabled Brill to capture images of any location on the mountain, zoom in, cut, and import into PowerPoint. He then added whatever overlays were necessary to identify the key features of the scene of the accident.

Brill realized that advanced mapping capabilities had potential benefits that extended far beyond accident investigation. For example, they can pinpoint the exact location of buried utilities for the construction and maintenance departments. Brill sold Keystone’s senior management on the concept, bought some GIS (Geographical Information Systems) software and, using highly sophisticated GPS units, started precisely mapping thousands of locations on Keystone’s property. The result was a comprehensive geographical database of the resort. (For related story, see “The Digital Mountain,” SAM May 2005.)

“The information has a wide range of uses,” Brill says. “One of them is orthophotography. You can import photographs and stretch them out so they’re geographically correct. And then you can overlay digital elevation data.”

By doing so, he can create accurate three-dimensional images out of two-dimensional ones, allowing him to animate fly-bys of a trail and an accident scene. Such images can closely reproduce what a person involved in an accident saw prior to the incident. For example, he can recreate the point of view of a skier riding up and unloading a chairlift.

Brill was recently involved in a case in which the resort had accurate baseline survey data. By utilizing that data in conjunction with measurements and images gathered by the AI team, he was able to show a jury what the plaintiff saw as he approached the halfpipe at the center of the case. (You can see several examples of Brill’s work at www.saminfo.com, or contact him at mountaingraphix@hotmail.com.)

Given the realities of day-to-day patrolling, much of Brill’s current work is beyond the scope of most ski patrol and risk management departments. But resorts can adopt many of his techniques to create their own AI reports, and Brill works with individual ski patrols to train them in the basics of using computers for basic AI work.

“At the simplest level,” he says, “all you really need is a computer with PowerPoint. We teach investigation teams how to use it to its potential. It takes some time, because most people use the application simply to create and show slides. We show them how to make maps, diagrams and photo logs.”

The result is investigation reports that look more professional and take less time to complete. File storage is simplified, and anyone who needs to see the report can receive it by e-mail.

Is that worth the trouble of using Brill’s approach? Perhaps. Brill believes that a well-executed accident investigation can mean “the difference between a $1,500 settlement and a $15 million one.” If that seems a bit hyperbolic, Rietz points out that “the more time you spend doing a good job, the easier it is to defend.”

Just remember to stick to the facts.

“If the area has some responsibility, so be it—we’re looking for the truth here,” says Rietz. “We don’t want to color anything. We want accurate information, in order to have the best course going forward.”