Laura Hollister, a citizen of Michigan, was severely burned when the shirt that she was wearing ignited upon contact with a hot electric burner on her apartment stove. She brought a products-liabilitylawsuit based upon an alleged design defect against Dayton-Hudson Corporation, the Minnesota-based owner of the department store where the shirt was purchased. Finding that Hollister had failed to establish a prima facie case of design defect under Michigan's "risk-utility test," the district court granted Dayton-Hudson's motion for summary judgment.

The district court concluded that Hollister had failed to present any evidence of a reasonable alternative design that would have reduced the risk of the type of injury that she had suffered, as required under the risk-utility test. She also failed to present any evidence as to the cost of reasonable alternatives, or their effect on the product as a whole. Because these elements are required by Michigan caselaw to establish a prima facie case of design defect, we AFFIRM the district court's decision.

In November of 1995, Hollister was a business student at Northwestern University in Evanston, Illinois. On November 4, 1995, she attended a business-school party with her friend Jerome Joliet. She later returned to her apartment accompanied by another friend, Diarmuid O'Connell, at approximately 1:45 a.m. According to O'Connell, Hollister was intoxicated when the two left the party. O'Connell left Hollister's apartment at approximately 2:10 a.m.

Hollister has no memory of subsequent events. The next thing that she can recall is seeing herself in the mirror at approximately 9:30 a.m. on November 5. Realizing that she was injured, Hollister phoned her parents in Michigan. According to Hollister's parents, she repeated the words "fire, burner, pasta." Hollister's parents asked her for the number of a friend and she supplied O'Connell's. After leaving a message on O'Connell's machine, Hollister's father asked his daughter for another number, and she supplied Joliet's. Mr. Hollister then left a message on Joliet's machine. Joliet arrived at Hollister's apartment at about 10:00 a.m. that day. Finding that Hollister was severely burned, Joliet called 911. Evanston Fire Department paramedics came immediately and treated Hollister.

A paramedic stated that the right front and rear burners of Hollister's stove were glowing red when he arrived. There was a bowl of cooked pasta in the sink, and a pot sitting on the stove between the burners. The fire department report concludes that the fire began in the kitchen. At the request of the fire department, the Bureau of Alcohol, Tobacco, and Firearms assisted in investigating the accident. The BATF report also indicates that the fire began in the kitchen. It concludes that the accident most likely occurred when Hollister reached for something in the cabinet above the stove (the door to which was open) and her shirt-tail, which was hanging loose, brushed against the burner and ignited. Hollister next apparently attempted to smother the flames on the counter, where burned cloth was found. The report states that there was evidence that she then attempted to extinguish the flames with water from the bathtub. There was evidence that after doing so, Hollister rested on her bed, where more burned clothing and bodily fluids were found. Small remnants of Hollister's shirt, a brown and black plaid button-down, were found. Hollister's mother stated that she had purchased this shirt for Hollister at a Hudson's department store six years earlier. Hollister was also wearing a T-shirt and bra under the large plaid shirt at the time of the accident. The report concludes that "the cause of this fire should be classified as accidental caused by ignition of the victim's clothing by the stove burner while she was cooking."

Hollister was brought to Evanston Hospital, where she was treated for third-degree burns over 55% of her body. She was later transferred to Loyola Hospital in Chicago, where she stayed until December21, 1995. After treatment at Loyola, Hollister was transferred to the University of Michigan Medical Center in Ann Arbor, Michigan, where she remained until April 17, 1996.

As a result of her burns, Hollister has undergone comprehensive skin grafting to most of her upper body, as well as plastic and reconstructive surgery. She remains profoundly disfigured. Her medical expenses at the time of the district court's ruling in May of 1998 amounted to approximately $980,000.

On March 27, 1996, Hollister's parents filed this products-liability lawsuit in the Circuit Court of Wayne County, Michigan. Dayton-Hudson (the parent company of the department store where the shirt was purchased), Ralph Lauren (which Hollister's mother originally believed to be the manufacturer of the shirt), Banana Republic (the retailer of the T-shirt), Victoria's Secret (the retailer of the bra), and General Electric (the manufacturer of the stove) were named as defendants. The complaint alleged that the shirt was defectively designed because it was dangerously flammable and because it did not provide a warning of its propensity to burn. On July 11, 1996, the case was removed to the United States District Court for the Eastern District of Michigan on the basis of diversity of citizenship. Pursuant to a July 31, 1997 motion, Hollister was substituted for her parents as the plaintiff.

On August 1, 1997, the district court held a status conference with all of the parties. At that time, the court noted that in the 16 months that the case had been in existence, a "whole lot of nothing" had taken place. At that point, the Hollisters had arranged for no expert witnesses other than a consultant who was going to conduct tests of various fabrics. The court directed Hollister to secure and disclose her experts, pursuant to Rule 26 of the Federal Rules of Civil Procedure, by September 1, 1997. A discovery deadline of October 31, 1997 was also set.

On September 1, 1997, Hollister submitted reports pursuant to Rule 26 from the following four experts: David Hall (textile expert), Edmund Knight (expert on cause and origin of the fire), Anna Dutka (economic damages expert), and Alan Hedge (stove design expert). Dr. Hall's report stated that he was still looking for "exemplar" fabric identical to the rayon used in the shirt in question, and that in his opinion the fabric was dangerously flammable. The report identifies the fabric as 100% rayon, loosely woven with 1.5 denier threads.

Dr. Hall offered no opinion as to the feasibility of using a different fabric to construct a similar shirt, and acknowledged that he had no expertise in the use of fabrics in clothing. He initially testified that the flammability test set forth in 16 C.F.R. 1610 determines whether a fabric is "unreasonably dangerous." This section's purpose is "to reduce danger of injury and loss of life by providing, on a national basis, standard methods of testing and rating the flammability of textiles and textile products for clothing use, thereby discouraging the use of any dangerously flammable clothing textiles." 16 C.F.R. 1610.1. Based upon their products' passing this test, Banana Republic and Victoria's Secret were eventually dismissed as defendants. Although Dr. Hall never tested the shirt remnants or "exemplar" fabric pursuant to 16 C.F.R. 1610, Dayton-Hudson's expert did. The fabric passed the test. Despite this fact, Dayton-Hudson remained as a defendant. By October of 1997, however, Hollister had dismissed all of the other remaining parties.

In his deposition on October 31, 1997, Dayton-Hudson's cause expert, John Campbell, acknowledged that he had located an "exemplar" shirt composed of fabric identical to that used in the shirt involved in the accident. Hollister's counsel subsequently purchased identical shirts to the one that Campbell had identified, and gavethem to Dr. Hall for examination and testing.

Dr. Hall determined that the characteristics of the exemplar fabric were substantially identical to the shirt that Hollister had been wearing at the time of the accident. He then conducted a test comparing the exemplar fabric with 14 other fabrics. The test utilized a stove-top electric burner set at 1100-1160 degrees, and involved sweeping 3.5 by 10-inch strips of the various fabrics across the burner. Hall then timed the rate of ignition and burning, if any. The exemplar fabric ignited immediately, and burned completely within six seconds. Eleven of the 14 non-exemplar samples failed to ignite. The three samples that did ignite were another 100% rayon sample, a rayon-polyester blend (both of which took about 12 seconds to burn completely), and a piece of newspaper (which burned in 4 seconds).

At the close of discovery, Dayton-Hudson filed three motions for summary judgment. One of the motions focused upon the requirement that a plaintiff such as Hollister prove the effectiveness of a proposed alternative design. In response, Hollister acknowledged that she would not be calling an expert witness on the effectiveness of an alternative design. She claimed that this was a question of fact for the jury that did not require expert testimony. The district court subsequently held a hearing on January 5, 1998 to clarify Hollister's position, stating as follows:

THE COURT: --you say that this is so clear, the risk of severe injury from a highly flammable fabric is so manifest, that no expert testimony is needed. That-- that really made me open my eyes. You mean we can simply impanel a jury, and you can put on your proof without any testimony to show this?

. . .

MR. HANDLER: We mean it.

The district court then presented the parties with a list of six requirements necessary to establish a prima facie case of design defect under Michigan law.

After briefing by both parties, the district court granted Dayton-Hudson's motion for summary judgment on May 12, 1998. Hollister timely filed her notice of appeal to this court.

We review de novo the district court's grant of summary judgment. See, e.g., Smith v. Ameritech, 129 F.3d 857, 863 (6th Cir. 1997). Summary judgment is appropriate when there are no issues of material fact in dispute and the moving party is entitled to judgment as a matter of law. See Fed. R. Civ. P. 56(c). In deciding a motion for summary judgment, the court must view the evidence and draw all reasonable inferences in favor of the non-moving party. See Matsushita Elec. Indus. Co., Ltd. v. Zenith Radio Corp., 475 U.S. 574, 587 (1986). The judge is not "to weigh the evidence and determine the truth of the matter but to determine whether there is a genuine issue for trial." Anderson v. Liberty Lobby, Inc., 477 U.S. 242, 249 (1986). A genuine issue for trial is presented when there is sufficient "evidence on which the jury could reasonably find for the plaintiff." Id. at 252.

B. The district court's six-part analysis comports with Michigan law in substance, if not in form

1. The elements of a design-defect case under Michigan law

The crux of Hollister's claim is that the shirt that she wore was defective because it was constructed of a fabric that was too highly flammable to be safe. In other words, Hollister argues that the dangerousness of the shirt is attributable to the "intentional design decisions of the manufacturer." See Prentis v. Yale Mfg. Corp., 365 N.W.2d 176, 182 (Mich. 1984)(distinguishing design defects from manufacturing defects).

In Owens v. Allis-Chalmers Corp., 326 N.W.2d 372, 378-79 (Mich. 1982), the Michigan Supreme Court held that a plaintiff in a design-defect case must present evidence of the magnitude of risks involved, and the utility and relative safety of proposed alternatives, in order to overcome the defendant's motion for summary judgment. The forklift operator in Owens was killed when the forklift he was driving overturned and pinned him beneath it. His wife argued that the forklift was defective because it lacked seatbelts. In granting a directed verdict for the defendant, the court stated that the plaintiff could not prevail because she had not shown that the presence of a seatbelt would have rendered the forklift safer. See id.

In Prentis, the Michigan Supreme Court built upon Owens and explicitly adopted a so-called "risk-utility" test for determining a manufacturer's liability. See Prentis, 365 N.W.2d at 186 ("Thus we adopt, forthrightly, a pure negligence, risk-utility test in products liability actions . . . where liability is predicated upon defective design."). The risk-utility analysis focuses upon whether a manufacturer "would be judged negligent if it had known of the product's dangerous design at the time it was marketed." Id. at 183.

In Reeves v. Cincinnati, Inc., 439 N.W.2d 326 (Mich. App. 1989), a worker whose hand was crushed in a power press that unexpectedly cycled sued the press's manufacturer on a design-defect theory. Reeves demonstrated that sudden cycling of power presses was common and that injury resulting from such cycling could occur. He further presented evidence that a guard installed in front of the press would prevent injuries such as the one he had incurred, and that such guards only cost about $1 each. On the other hand, he did not provide evidence as to any costs that might be associated with redesigning the presses to accommodate the guards, or the exact statistical likelihood of injury. See id. at 329-30.

In reversing a directed verdict for the manufacturer, the Michigan Court of Appeals stated that the plaintiff had presented sufficient evidence to raise a question of fact for the jury as to the reasonableness of the power press's design. See id. at 330. The Reeves court succinctly articulated the following elements of a prima facie design-defect case under Michigan law:

[A] prima facie case of a design defect premised upon the omission of a safety device requires first a showing of the magnitude of foreseeable risks, including the likelihood of occurrence of the type of accident precipitating the need for the safety device and the severity of injuries sustainable from such an accident. It secondly requires a showing of alternative safety devices and whether those devices would have been effective as a reasonable means of minimizing the foreseeable risk of danger. This latter showing may entail an evaluation of the alternative design in terms of its additional utility as a safety measure and its trade-offs against the costs and effective use of the product.

Id. at 329 (citing Prentis).

This court has previously applied the Owens, Prentis, and Reeves analysis in a design defect case under Michigan law. See Zettle v. Handy Mfg. Co., 998 F.2d 358, 360 (6th Cir. 1993) (holding that the plaintiff had failed to present sufficient evidence concerning the effectiveness of a proposed alternative design in a products liability action against the manufacturer of a power washer).

In the present case, Hollister has sued the distributor of the shirt, not its manufacturer. The manufacturer of the shirt remains unknown. A distributor's liability is the same as that of the manufacturer in a design-defect case. See Gregory v. Cincinnati, Inc., 538 N.W.2d 325, 339 (Mich. 1995).

2.The district court's six-part prima facie case

Citing Owens, Prentis, and Reeves, the district court characterized the Michigan law on design defect as requiring Hollister to produce evidence that shows:

(1). that the severity of the injury was foreseeable by the manufacturer;

(2). that the likelihood of occurrence of her injury was foreseeable by the manufacturer at the time of distribution of the product;

(3). that there was a reasonable alternative design available;

(4). that the available alternative design was practicable;

(5). that the available and practicable reasonable alternative design would have reduced the foreseeable risk of harm posed by defendant's product; and

(6). that omission of the available and practicable reasonable alternative design rendered defendant's product not reasonably safe.

Hollister argues on appeal that this test held her to a higher burden than the risk-utility test employed by the Michigan courts. In particular, she states that Michigan's analysis is, in fact, no more than a complicated way of expressing the concept of "reasonable care." Although it is true that the Michigan risk-utility test is a way of determining whether manufacturers exercise reasonable care in their design decisions, it is also true that Hollister had to meet the requirements set forth in Owens, Prentis, and Reeves in order to survive summary judgment.

Our review of the district court's test and its opinion leads us to the conclusion that the district court laid out the same elements of a prima facie design-defect case that the Michigan courts require. Steps one and two of the district court's test--the foreseeability of serious injury and the likelihood that injury would occur-- simply restate the following language in Reeves: "a showing of . . . the likelihood of occurrence of the type of accident . . . and the severity of injuries sustainable from such an accident." 439 N.W.2d at 329.

Steps three, four, and five required Hollister to present evidence that a reasonable design alternative was available, that it was practicable, and that it would have reduced the risk of the accident at issue in the case. These requirements simply reiterate the second half of the Reeves test, requiring "a showing of alternative safety devices and whether those devices would have been effective as a reasonable means of minimizing the foreseeable risk of danger." Id.

Step six, stating that the product must be unreasonably dangerous without the alternative design that was available and practicable, expresses the requirement in Owens that the absence of the alternative design be the cause of the injuries. See Owens, 326 N.W.2d at 379 (holding that the plaintiff had failed to show the reasonableness of requiring a seatbelt in part because there was no evidence that it was practicable for forklift drivers to use a seatbelt in performing their tasks). We therefore conclude that the district court applied the correct test in adjudicating Dayton-Hudson's motion for summary judgment.

C. Hollister failed to establish a prima facie case of design defect

Hollister's case basically rests upon the test conducted by Dr. Hall in which the exemplar fabric and 14 other fabric samples were dragged across a burner set at between 1106 and 1160 degrees. Dayton-Hudson points out that a General Electric engineer stated that the stove in Hollister's apartment would burn at between 1500 and 1800 degrees when on the "high" setting (where it was set at the time the paramedics entered Hollister's apartment). Hollister states that Dayton-Hudson'sexpert was talking about new stoves, and that a used stove such as that in Hollister's apartment might burn at a lower temperature. She provides no support for this assertion, however, nor any justification for failing to conduct the test at the higher temperature.

In the absence of conditions more like those in Hollister's apartment, the utility of Dr. Hall's test is questionable. Because it was unknown at the close of discovery whether the alternative fabrics would burn on contact with a much hotter burner, it is impossible to say whether such fabrics would have reduced the risk of the accident at issue here. According to Dayton-Hudson, for example, wool has an ignition temperature of over 1,100 degrees, which is right at the upper limit of Dr. Hall's test. As the district court noted, Hollister's own expert "stated that he could not articulate the exact effect on flammability of the changes Hollister claims would make the shirt safer." Although he asserted that changing the denier thread thickness of the fabric would have made it less flammable, Dr. Hall said that he would need a "crystal ball" to predict the extent of the change on flammability.

Furthermore, the district court correctly concluded that Hollister had presented no evidence whatsoever as to the availability of alternative fabrics when the shirt was manufactured, the cost of manufacturing the shirt with such fabrics, or the effect of a fabric change upon the wearability, durability, or appearance of the fabric. Hollister's failure to submit such evidence is similar to the situation in Owens, where the plaintiff offered no information as to the practicability of a seatbelt on a forklift. See 326 N.W.2d at 379.

In fact, Dr. Hall never presented a "proposed alternative design" at all, because he never specified the construction details of an alternative shirt. His only recommendation was that the weight of the fabric be heavier. Therefore, as a practical matter, there is no proposed alternative to compare to the shirt that Hollister wore at the time of the accident.

The lack of a specific proposed alternative is particularly damaging to Hollister's case in light of Dr. Hall's test results. Fabric B in the test, a 100% rayon sample that was heavier than the exemplar fabric, burned as quickly as the exemplar. Fabric C, a second 100% rayon sample that was heavier than the exemplar but lighter than fabric B, did not even ignite. Without exact weight specifications for the proposed alternative shirt, it is therefore impossible to determine the effect that an alternative fabric would have upon flammability.

Because Hollister failed to establish a nexus between the manufacturer's choice of fabric and the injury, or the practicability of an alternative fabric, she was unable to demonstrate that the shirt was "not reasonably safe." Her burden was all the greater because, as previously stated, the exemplar fabric actually passed the federal flammability test set forth in 16 C.F.R. 1610. She therefore failed to make out a prima facie case of design defect under Michigan law.

D. The district court's evaluation of other factors was inconsistent with Michigan law

Rather than rest its decision solely on Hollister's failure to meet requirements 3 through 6 of Michigan's risk-utility test, the district court also ruled against her on the alternate grounds of severity (requirement 1), foreseeability (requirement 2), the "open and obvious danger" doctrine (one of Dayton-Hudson's affirmative defenses), and the alleged misuse of the shirt (another Dayton-Hudson affirmative defense). Because the district court's opinion was published, see 5 F. Supp. 2d 530 (E.D. Mich 1998), and because we find that its rulings on these alternate grounds were erroneous under the circumstances of the case before us, we discuss each of them below.

In Part VI.A.1. of its opinion, the district court erred in concluding, as a matter of law, that the severity of Hollister's injuries from a burning shirt was not foreseeable. A 1985 report of the Consumer Products Safety Commission ("CPSC") states that "[o]ver one third of all clothing-related burn victims were hospitalized. This fact becomes dramatic when compared to the 4 percent hospitalization rate for all consumer product-related injuries . . . and the 8 percent reported for all burn injuries." Furthermore, Hollister presented evidence that the fabric in question would ignite on contact with an electric burner and be consumed within seconds. In addition to the test comparing strips of fabric, her videotape shows the flames shooting above the shoulders of an exemplar shirt that her expert burned in its entirety.

In Zettle, a panel of this court noted that once the plaintiff had demonstrated a risk of electrocution in the use of a pressure washer, passing the "severity test" was easier because a likely result of severe electric shock is death. See 998 F.2d at 363. Similarly, incurring severe burns is a likely result of being engulfed in flames. Hollister's demonstration that the shirt would burn quickly, coupled with the CPSC's report on hospitalization from clothing burns, creates a question of fact for the jury on the foreseeability of severity.

In Part VI.A.2. of its opinion, the district court erred in concluding, as a matter of law, that the likelihood of Hollister's injury was not foreseeable. Hollister presented the 1985 CPSC report setting forth statistics on injuries resulting from apparel catching fire. The report specifically indicates that kitchen ranges are a common ignition source in accidents involving burning apparel. In fact, the report refers to precisely the type of accident in which Hollister was involved:

Kitchen ranges were the second most common ignition source. Ignition occurred most frequently when adults were leaning against or reaching across a range while wearing shirts/blouses, when children were climbing on or playing with ranges while wearing pajamas, and when elderly women were cooking while wearing robes or housecoats.

Also included in the report are statistics showing the number of injuries, by age group, that result from shirts' igniting on stove tops. In concluding that Hollister had not presented sufficient evidence regarding the magnitude of the risk involved, the district court noted that the CPSC report was from 1985, four to five years prior to the purchase of the shirt. It did not explain, however, why the date of the report invalidates it as evidence. The district court then analyzed the statistics contained therein and concluded that there would only be 123 injuries per year in Hollister's age group that might match the severity of the injury that she incurred. It concluded, without citation or explanation, that "these are small numbers indeed." The district court thus judged the sufficiency of the evidence presented, which is properly a question for the jury, not the court. See Zettle, 998 F.2d at 360.

In Part VI.C. of its opinion, the district court opines that Hollister should not prevail in a case such as this one because the danger from clothing touching a hot stove is "open and obvious." We find this to be a misapplication of the "open and obvious danger" doctrine, which was articulated in Fisher v. Johnson Milk Co., Inc., 174 N.W.2d 752 (Mich. 1970). In that case, the plaintiff slipped on a patch of ice and dropped a wire carrier full of glass milk bottles. The bottles broke, injuring him. The plaintiff claimed that the manufacturer should have placed a bottom on the carrier, which would have prevented glass fragments from scattering when the case was dropped. See id. at 752-53. Inaffirming the grant of summary judgment for the defendant, the Michigan Supreme Court stated that the danger of bottles breaking when dropped was obvious to everyone, and that there was no duty to construct the carrier differently or to warn consumers. See id.

The "open and obvious danger" doctrine set forth in Johnson Milk is not applicable in the present case because the relative flammability of clothing is not obvious to the average consumer. Although the danger of a shirt's igniting when placed in contact with a hot stove is obvious, the relative flammability of one shirt versus another is not. Hollister alleged that the exemplar fabric burned explosively, like newspaper, but that other fabrics did not. Had she established a prima facie case, she would not have been hindered by the "open and obvious danger" doctrine.

The district court further opined in Part VI.C. that Hollister's own actions ought to preclude her from relief, based on Hollister's alleged misuse of the clothing. Manufacturers, however, have a duty to warn consumers of the dangers associated with foreseeable misuses of the products they produce. See Allen v. Owens-Corning Fiberglass Corp., 571 N.W.2d 530, 534-35 (Mich. App. 1997). Hollister's failure-to-warn claim explicitly states that had she known that the consequences of a mishap such as hers would be so high, she would not have worn the shirt. As noted above, the CPSC study indicates that accidentally brushing a stove burner as Hollister did is a foreseeable misuse of clothing.

Furthermore, the district court held, without support, that Hollister was a "careless user" of the clothing. Although the manufacturer is not liable for every possible misuse of its product, see Owens, 326 N.W.2d at 379, the district court was not entitled on the record before us to conclude as a matter of law that Hollister's alleged misuse was so grievous as to deprive her of the right to pursue a design-defect lawsuit. We therefore conclude that Hollister's alleged misuse and carelessness were not proper issues to be decided by summary judgment in the case before us.

In sum, we AFFIRM the decision of the district court based upon Hollister's failure to present evidence that a reasonable alternative design was available, that it was practicable, that it would have reduced the risk of the accident at issue, and that the absence of the alternative design was the cause of her injuries. On the other hand, we find that the district court's rulings on severity (Part VI.A.1.), foreseeability (Part VI.A.2.), the "open and obvious danger" doctrine (Part VI.C.), and misuse (Part VI.C.) are not consistent with Michigan law.