Plaintiff, Tari Marshall, filed suit against her former employer, American Hospital Association ("AHA"), alleging that AHA terminated her employment on the basis of pregnancy in violation of the Pregnancy Discrimination Act, 42 U.S.C. § 2000e(k), an amendment to the Civil Rights Act of 1964 ("Title VII"), 42 U.S.C. § 2000e et seq. After discovery was completed, AHA filed a motion for summary judgment. The district court granted the motion and entered summary judgment in favor of AHA. Marshall appeals.

In August 1994, AHA placed an employment advertisement in the Chicago Tribune, soliciting applications for an Associate Director position in its American Society for Health Care Marketing and Public Relations division ("Society"). The advertisement stated that applicants should have a bachelor's degree in journalism, marketing or communications and seven to eight years of health care experience in public relations, marketing or communications. Tari Marshall, who had extensive experience in public relations and marketing, saw the ad. Although she did not have health care experience, Marshall sent the Society her resume. After receiving the resume, in October 1994, AHA contacted Marshall to set up an interview.

On November 2, 1994, Marshall had her first interview with AHA. Lauren Barnett, Executive Director for the Society, interviewed Marshall and described the Associate Director's responsibilities, which included the organization of the Society's government relations symposium in January and its annual education conference in September. Barnett stressed the importance of the September conference, which provided 40% of the Society's annual revenue and was the Society's premier event of the year. Barnett told Marshall that the conference represented a large time commitment for the Associate Director, whose tasks for the conference included soliciting speakers, assisting the speakers with their preparations, coordinating the program, and attending the conference. Marshall also met with Sidney Cristol, another Associate Director at the Society, who told Marshall that she enjoyed working for the Society, but that the atmosphere was "very intense, that it was very hard work, that there was a lot to do," and "that because it was a small staff, ... everybody was kept very, very busy." Marshall knew that she was pregnant at this time and that the baby was due at the end of June 1995, but she did not inform anyone at the Society of her condition.

Several weeks later, Marshall was called for a second interview with the Society. At this second interview, Marshall and Barnett discussed Marshall's need to become more knowledgeable about health care issues. Marshall told Barnett that she was "a pretty quick study" and that she was prepared to do a lot of independent work to acquire the health care knowledge that she needed; however, both women knew that Marshall would need some time to familiarize herself with the health care field.

Barnett contacted Marshall's references, and one of them told her that Marshall had recently had a miscarriage and was trying to become pregnant. Barnett and Cristol discussed the need to fill the Associate Director position and Marshall's lack of health care experience. Cristol suggested looking at other candidates, but Barnett told Cristol that she believed that she could teach health care to the right individual. Barnett made the decision to hire Marshall on November 23, 1994. Marshall began working for AHA on December 12, 1994, with the understanding that her first ninety days of employment were considered a probationary period.

During Marshall's first week of work, Barnett apologized to Marshall for not having spent much time with her, but told her that everything seemed to be going well. On December 27, 1994, approximately two weeks after beginning work at the Society, Marshall informed Barnett that she was pregnant and that the baby was due at the end of June 1995. Marshall told Barnett that she was planning on working until the baby was born and anticipated taking an eight-week maternity leave beginning after the birth.

Barnett expressed concern about Marshall's absence during the months just before the September conference, the busiest time of the Society's year, and told Marshall that her absence would cause extra work for Barnett and Cristol. Barnett doubted that the Society could afford to pay Marshall's salary during her leave and to hire a temporary employee to fill the vacancy. Barnett told Marshall that the other Associate Director candidate she had been considering did not have "this issue." Marshall assured Barnett that she would return to work at the end of her maternity leave. Barnett responded by questioning Marshall about her child care and financial arrangements. Barnett stated that she had learned that Marshall was attempting to get pregnant from one of her references, but that she had no idea Marshall would succeed so soon. Marshall offered to take unpaid leave and to install office equipment in her home to allow her to continue working during her leave. Marshall wanted to tell the rest of the staff about her pregnancy so they could begin to prepare in advance for the September conference; however, Barnett requested that Marshall wait to tell anyone other than Cristol of her pregnancy. Barnett said that the news of the pregnancy might cause the rest of the staff to become upset or to panic about its effect on the educational conference and suggested that Marshall give the others a chance to get to know her and to work with her before telling them.

Following her meeting with Barnett, Marshall told Cristol that she was pregnant and that Barnett had expressed concerns. Cristol congratulated Marshall and urged her not to pay any attention to Barnett. Cristol said that the Society would be fine in Marshall's absence, since the Society had been without an Associate Director during the same time period the previous year. Cristol told Marshall that Barnett was unsympathetic toward her employees' family needs because Barnett was unable to have children of her own.

Shortly after Marshall disclosed her pregnancy to Barnett, Barnett asked her to write a membership recruitment letter. Normally, this type of assignment was performed by Cristol, not Marshall. Marshall alleges that Barnett gave her this assignment outside of her field to set her up for failure because of her pregnancy. AHA contends that Cristol had other tasks, that the letter needed to be sent out, and that Marshall should have been capable of writing such a letter considering her public relations and marketing background. Marshall asserts that the recruitment letter was the only new substantive assignment given to her after her pregnancy announcement. In any event, Barnett was displeased with the letter because she believed that it merely repeated information set forth in the Society's membership brochure and did not differentiate the Society from its competitors. Barnett also asked Marshall to write a letter to the CEO of a health care corporation inviting him to speak at the September conference. Barnett ended up writing the letter herself because she thought that Marshall's draft was unsuitable.

Marshall contends that Barnett prohibited her from becoming substantially involved in the January Government Relations Seminar. It is undisputed that most of the work for the seminar had been completed before Marshall was hired. Marshall was assigned to sit at a registration table during the seminar and was asked to stuff envelopes along with other staff members including Barnett and Cristol. Marshall was not allowed to sit in on the sessions at the seminar; however, because Society staff members were busy coordinating the seminar, the presentations were taped and the staff was allowed to listen to the tapes after the seminar. Additionally, Marshall testified that after she made her pregnancy announcement, "there were just a lot of times, too, when [she] just could tell [Barnett's] anger."

On January 19, 1995, Cristol wrote a memo to Barnett outlining her concerns about Marshall's attempts to procure speakers for the September conference and about Marshall's relationship with the other people involved in the project. By the end of January, Barnett had decided to terminate Marshall's employment with the Society. Barnett wrote a memo to the file dated January 31, 1995, outlining her decision to terminate Marshall based on Marshall's lack of health care experience which resulted in an unsatisfactory level of job performance. On February 6, 1995, about eight weeks after Marshall began working at AHA, Barnett told Marshall she was fired because her lack of health care background was causing problems with her work. Marshall responded that Barnett had never brought any problems to Marshall's attention and asked Barnett to provide specific examples of performance problems. Barnett pointed to the draft membership letter and the Call for Speakers for the September conference as problem areas.

After her termination, Marshall wrote a letter to the AHA Human Resources Department, asserting that Barnett had terminated her because she was pregnant and requesting a copy of her personnel file. In her letter, Marshall stated that Barnett had "repeatedly said that [Marshall's] leave would cause problems, including financial concerns for the Society." When Marshall received her personnel file from AHA, it did not include either the memo from Cristol to Barnett describing Marshall's poor relations with the other staff members or the January 31, 1995 file memo from Barnett outlining the influence of Marshall's lack of health care expertise on Barnett's decision to terminate her. Marshall's position at AHA was filled in May 1995.

We review the district court's grant of summary judgment de novo. Kennedy v. Schoenberg, Fisher & Newman, Ltd., 140 F.3d 716, 722 (7th Cir.1998) (citing Geier v. Medtronic, Inc., 99 F.3d 238, 240 (7th Cir.1996)), petition for cert. filed, 67 U.S.L.W. 3082 (U.S. July 2, 1998) (No. 98-66). We will affirm if the record, viewed in the light most favorable to the non-moving party, reveals that there is no genuine issue as to any material fact and the moving party is entitled to judgment as a matter of law. Fed.R.Civ.P. 56(c). "The mere existence of some alleged factual dispute between the parties will not defeat an otherwise properly supported motion for summary judgment; the requirement is that there be no genuine issue of material fact." Anderson v. Liberty Lobby, Inc., 477 U.S. 242, 247-48, 106 S. Ct. 2505, 91 L. Ed. 2d 202 (1986). We apply the summary judgment standard with particular care in employment discrimination cases, which often turn on the issues of intent and credibility. Kennedy, 140 F.3d at 722.

In 1978, Congress amended Title VII of the Civil Rights Act to prohibit discrimination on the basis of pregnancy. Kennedy, 140 F.3d at 722 (citing Geier, 99 F.3d at 241). This amendment, known as the Pregnancy Discrimination Act ("PDA"), provides that "[w]omen affected by pregnancy, childbirth or related medical conditions shall be treated the same for all employment-related purposes ... as other persons not so affected but similar in their ability or inability to work." 42 U.S.C. § 2000e(k). The PDA does not require employers to offer maternity leave or other special assistance to pregnant employees; however, it does require employers to treat a pregnant employee as well as that employee would have been treated had she not been pregnant. Kennedy, 140 F.3d at 722 (citation omitted); Troupe v. May Dep't Stores, 20 F.3d 734, 738 (7th Cir.1994).

We must determine whether Marshall presented a question of fact as to whether AHA treated her less favorably because of her pregnancy. There are two ways in which a plaintiff may establish a case of pregnancy discrimination in order to avoid summary judgment. First, the plaintiff may use the indirect, burden-shifting approach set out in McDonnell-Douglas Corp. v. Green, 411 U.S. 792, 93 S. Ct. 1817, 36 L. Ed. 2d 668 (1973). Kennedy, 140 F.3d at 723. Alternatively, the plaintiff may proceed under the "mixed motives" or direct method, in which case the plaintiff may avoid summary judgment by producing sufficient evidence, either direct or circumstantial, to create a triable issue as to whether pregnancy was a motivating factor in her discharge. Id. at 722 (citing Geier, 99 F.3d at 241); Gleason v. Mesirow Financial, Inc., 118 F.3d 1134, 1140 (7th Cir.1997). Marshall does not attempt to utilize the McDonnell-Douglas burden-shifting method; rather, she contests the district court's determination that she could not meet her burden under the direct approach. Under this approach, a plaintiff may establish discriminatory motivation through either " 'direct evidence, e.g., an acknowledgment on the part of the employer of discriminatory intent, or--as is more usually the case--by relying on circumstantial evidence, e.g., ambiguous statements or suspicious timing.' " Id. (quoting Geier, 99 F.3d at 241). Marshall admittedly has no direct evidence of discrimination, so she attempts to meet her burden with circumstantial evidence.

In Troupe, this court noted that there are three types of circumstantial evidence under the direct approach. Each type may be sufficient by itself to support a judgment for the plaintiff, or the three types may be used together. Troupe, 20 F.3d at 736. The first type consists of "suspicious timing, ambiguous statements oral or written, behavior toward or comments directed at other employees in the protected group, and other bits and pieces from which an inference of discriminatory intent might be drawn." Id. The second is a showing that other, similarly situated, non-pregnant employees received systematically better treatment. Id. The third is evidence that the plaintiff was qualified for the job in question but passed over in favor of a person not having the forbidden characteristic and that the employer's stated reason for its decision is "unworthy of belief, a mere pretext for discrimination." Id. See also Kennedy, 140 F.3d at 725 (citing Piraino v. International Orientation Resources, 84 F.3d 270, 274 (7th Cir.1996)).1 In order to avoid summary judgment, Marshall must produce evidence from which a rational trier of fact could reasonably infer that AHA fired her because she was pregnant. Id. at 737.

Marshall structures her argument under the first type of circumstantial evidence outlined in Troupe, relying on "suspicious timing, ambiguous statements oral or written, behavior toward or comments directed at other employees in the protected group, and other bits and pieces from which an inference of discriminatory intent might be drawn." First, Marshall contends that the record contains evidence of suspicious timing sufficient to avoid summary judgment. Specifically, Marshall alleges that she received no new assignments after her pregnancy announcement, that all of Barnett's criticisms of Marshall's work came after the announcement, and that Barnett was regularly angry with her only after the announcement. These allegations involve what Marshall perceives as a sudden change in Marshall and Barnett's relationship beginning from the time Marshall informed Barnett of her pregnancy. A sudden change in attitude upon disclosure that the plaintiff is member of a protected class may raise an inference of discrimination if the plaintiff can establish that the only intervening event was the disclosure. See Leffel v. Valley Financial Services, 113 F.3d 787, 794 (7th Cir.1997), cert. denied, --- U.S. ----, 118 S. Ct. 416, 139 L. Ed. 2d 318 (1997) ("A bank officer who is praised and given wide discretion in the performance of her duties one day but is confronted with a laundry list of manufactured criticisms and confining performance standards the next may be able to raise an inference that discrimination is afoot by establishing that the only intervening event was the disclosure that she is disabled."). In the present case, Marshall cannot meet this burden. It is undisputed that at the same time that Marshall disclosed her pregnancy she also informed Barnett that she was planning on being absent from work for two months during the busiest part of the Society's year. The PDA "requires the employer to ignore an employee's pregnancy, but ... not her absence from work, unless the employer overlooks the comparable absences of non-pregnant employees." Troupe, 20 F.3d at 738. Additionally, it is undisputed that at the time Barnett made the decision to hire Marshall she knew that Marshall was trying to become pregnant. What Barnett did not know, as she informed Marshall during the December 27 meeting, was that Marshall was planning on being absent from work for the two months prior to the annual educational conference. Barnett's alleged change in attitude toward Marshall does not raise the inference of pregnancy discrimination.

Marshall, relying on this court's retaliation cases, also contends that a jury could find that the fact that she was terminated within six weeks of her pregnancy announcement warranted an inference of discriminatory intent. However, in Hunt-Golliday v. Metropolitan Water, 104 F.3d 1004, 1011 (7th Cir.1997), this court held that in PDA cases, under a circumstantial evidence approach, timing alone does not create an inference of discrimination. A PDA plaintiff must show some connection between her pregnancy announcement and the unfavorable employment action. Id. at 1010-11 (holding that, despite the fact that plaintiff was suspended the day after announcing her pregnancy, plaintiff had failed to show any connection between her announcement and the suspension). Marshall fails to present any evidence of this necessary connection.

Additionally, Marshall points to Barnett's comments during the December 27 meeting and Cristol's comments following that meeting as ambiguous statements and actions supporting an inference of discrimination. In order for isolated comments to be probative of discrimination, the " 'comments must be contemporaneous with the discharge or causally related to the discharge decision-making process.' " Gleason, 118 F.3d at 1140 (quoting Geier, 99 F.3d at 242). Marshall concedes that the comments were not made contemporaneously with her termination. Therefore, Marshall must demonstrate a causal connection between the comments and Barnett's decision to terminate her employment. Marshall fails to do so; instead, she cites cases which hold that in indirect, McDonnell-Douglas proof cases, this causal connection need not be present. See, e.g., Huff, 122 F.3d at 383-86; Futrell v. J.I. Case, 38 F.3d 342, 347 (7th Cir.1994). However, this argument is futile because Marshall concedes that she is proceeding under the direct proof method as outlined in Troupe and not under the indirect, McDonnell-Douglas approach. Therefore, in the absence of a causal nexus between the statements and the termination decision, Marshall cannot rely on the statements as evidence of discrimination. See Cowan v. Glenbrook Sec. Services, Inc., 123 F.3d 438, 444 (7th Cir.1997); Geier, 99 F.3d at 241-42. Even if Marshall were proceeding in an indirect proof case, "remarks unrelated to the employment decision in question may not overcome summary judgment if they stand alone as evidence of the employer's discriminatory intent." Huff, 122 F.3d at 385 (citations omitted). Additionally, while Marshall contends that there is a genuine issue of fact as to whether Barnett's comments referred to Marshall's pregnancy or to her absence during the critical period of her first year at the Society, in her February 6, 1995 letter to Human Resources, Marshall herself states that when she informed Barnett of her pregnancy, Barnett "repeatedly said [her] leave would cause problems." (emphasis added).

Finally, Marshall argues that AHA's "shifting" reasons for the discharge support a finding of discrimination. Marshall contends that Barnett's explanation that she was firing Marshall because of her lack of health care knowledge was "suspicious" because Barnett knew of Marshall's lack of health care experience when she hired her. Additionally, Marshall asserts that AHA did not reveal that Marshall's termination was based in part on the fact that her maternity leave would interfere with the fall conference until after litigation had begun. This is not evidence from which an inference of discrimination can be drawn. Instead, this evidence fits under a pretext analysis as outlined under the third circumstantial evidence prong, yet Marshall acknowledges that she cannot proceed under a pretext approach. Under the pretext approach, disbelief of an employer's proffered reasons for termination must be coupled with a prima facie showing by the plaintiff in order to suffice as evidence of intentional discrimination. St. Mary's Honor Center v. Hicks, 509 U.S. 502, 511, 113 S. Ct. 2742, 125 L. Ed. 2d 407 (1993). See also Troupe, 20 F.3d at 736-37 (holding that because plaintiff could not make a prima facie showing that she was qualified for the job in question, "she could not raise an issue of pretext"). Marshall has not made the requisite prima facie showing.

Even assuming Marshall was able to establish a prima facie case, she would then have to prove that AHA's non-discriminatory reasons for firing her were pretextual. While Marshall tries to convince this court that "Barnett originally forgave Marshall her modest health care background," it is undisputed that Marshall had assured Barnett during the interview process that she was a quick study and would work independently to bring herself up to speed on health care issues. At the time she accepted the Associate Director position, Marshall was aware of Barnett's high expectations and of the highly intense atmosphere of the Society. At the end of Marshall's first week of work, Barnett apologized to Marshall for not spending more time with her and told Marshall that she seemed to be picking things up quite nicely; however, besides this comment, Marshall offers no evidence that she had indeed been able to develop the requisite health care knowledge at a level sufficient to meet Barnett's expectations. Furthermore, while Marshall argues that her absence from work was only raised as an issue after litigation began, it is undisputed that in her February 6, 1995 letter to Human Resources, Marshall stated that she believed that she was being fired due to her pregnancy because when she announced her pregnancy to Barnett, Barnett repeatedly told Marshall that her leave would cause problems for the Society. Clearly, Marshall was aware of the absence issue at the time of her termination.

As previously noted, the PDA "requires the employer to ignore an employee's pregnancy, but ... not her absence from work, unless the employer overlooks the comparable absences of non-pregnant employees." Troupe, 20 F.3d at 738. Marshall makes no showing that AHA terminated her because of her pregnancy rather than because she was planning an extended absence during the busiest time of her first year as an Associate Director. Marshall presents no evidence that a non-pregnant, probationary employee who was going to be on extended leave in the months immediately preceding the annual conference would not have been terminated. There is no evidence from which a rational trier of fact could infer that Marshall was the victim of pregnancy discrimination.

We agree with the district court's conclusion that the record evidence, viewed in the light most favorable to Marshall, does not constitute sufficient evidence of discrimination to allow Marshall to avoid summary judgment. Marshall fails to completely meet the necessary requirements of either the direct or indirect approaches in her effort to show the alleged discrimination. She attempts to borrow from one approach to make up for an evidentiary shortcoming in the other. The burden is met in neither. This sort of "mix and match" approach may work in some situations, but not here as each approach is distinct with its own requirements. The decision of the district court is AFFIRMED.