HIGGINSON, Circuit Judge:

A third-party administrator of medical benefits plans denied claims made on behalf of two patients who received treatment from the same medical provider. The district court found that (1) the plan administrator incorrectly interpreted the plans to deny the claims in a way that abused its discretion and (2) the administrator may be held liable for its wrongful denial. The district court also awarded attorneys' fees to the medical provider. We AFFIRM.

1. Facts and Proceedings

Christopher Evans suffered a cervical spine fracture that resulted in quadriplegia. Evans received treatment at LifeCare Management Services, LLC ("LifeCare") in Dallas, Texas for about two-and-a-half months before moving to a nursing home in July 2005. His medical bills totaled more than $171,000.

Bobby Wall suffered an acute stroke. Wall received treatment at a LifeCare facility in Shreveport, Louisiana for about two months before passing away in June 2007.1 His medical bills totaled more than $340,000.

Evans and Wall participated in similar medical benefits plans through Carter Chambers LLC ("Carter") and Bill and Ralph's Inc. ("BRI"), respectively. Evans was a Carter employee's dependent and a qualified participant of the Carter plan. Wall was BRI's employee and a qualified participant of the BRI plan. The plans listed Carter and BRI as administrators.

The plans limited reimbursements to "skilled nursing facilities" ("SNFs").2 The plans used identical language to define an "SNF":

The plans further provided in a final sentence that the term "skilled nursing facility" "also applies to charges incurred in a facility referring to itself as an extended care facility, convalescent nursing home, rehabilitation hospital, long-term acute care facility or any other similar nomenclature."

By contrast, the plans covered reimbursements to hospitals. The plans defined a "hospital" as:

Carter and BRI contracted with Insurance Management Administrators, Inc. ("IMA") to act as a third-party administrator ("TPA") of claims arising under the plans. The administration contracts between IMA and Carter and BRI outlined the scope of IMA's administrative duties. The contracts allowed IMA to "[p]rocess all claims presented for benefit under Plan, [...] audit claims processed by selected Insurance Carrier[s] to determine accuracy, distribute checks in payment of claims to employees or service providers, and provide an explanation of claim settlements to the Plan Participant and Plan Administrator." The contracts also specified that IMA's duties were "ministerial in nature" and to be "performed within the framework of policies, interpretations, rules, practices and procedures" established by the employers.

Referencing the plans' limits on SNF reimbursements, IMA refused to pay either Evans' or Wall's claims. Longtime IMA claim manager Alana Bennett denied Wall's claim by explaining that LifeCare did "not meet the definition of a hospital as defined in the plan" because LifeCare "is a rehab facility as defined in the plan," and the plan did "not have rehab benefits." Bennett denied Evans' claim by explaining that LifeCare was an SNF because it satisfied the first and sixth factors of the plan's seven-part SNF test: LifeCare helped Evans "convalesce from an injury" and was "licensed as a specialty hospital." Bennett also indicated to Evans that LifeCare qualified as an SNF under the plan's final sentence elaborating on SNFs because LifeCare was a long-term acute care facility ("LTAC").3

Bennett testified at her deposition that, even if a facility referred to itself as an LTAC, it would still have to meet each of the seven SNF factors to qualify as an SNF under the plan. She also testified that she denied LifeCare's claims because LifeCare did not meet the plans' seven-factor test.

After IMA denied Wall's and Evans' claims, LifeCare filed separate lawsuits against IMA, BRI, the BRI Plan, Carter, and the Carter Plan alleging that they wrongfully denied Wall's and Evans' claims under the Employee Retirement Income Security Act ("ERISA"). LifeCare also raised related state law claims.

The district court consolidated the cases. The parties filed motions for summary judgment. The district court granted summary judgment for IMA, BRI, the BRI Plan, Carter, and the Carter Plan on LifeCare's state law claims, but granted summary judgment for LifeCare on its ERISA claims. The district court found that IMA incorrectly interpreted the plans to categorize IMA as an SNF in a way that abused its discretion. The district court also found that LifeCare could maintain a claim against IMA as a TPA. The district court awarded LifeCare benefits payments in excess of $512,000 and attorneys' fees totaling more than $453,000.

IMA, BRI, the BRI Plan, Carter, and the Carter Plan (the "Appellants") raise three issues on appeal: (1) whether the district court erred in finding that IMA incorrectly interpreted the plans to deny payments to LifeCare in a way that abused its discretion; (2) whether the district court erred in finding IMA liable for its handling of LifeCare's claim; and (3) whether the district court erred in awarding attorneys' fees to LifeCare.

2. Standard of Review

This court reviews a grant of summary judgment de novo, applying the same standards as the district court. Trinity Universal Ins. Co. v. Emp'rs Mut. Cas. Co., 592 F.3d 687, 690 (5th Cir.2010). We therefore affirm the district court's grant of summary judgment "if, viewing the evidence in the light most favorable to the non-moving party, there is no genuine dispute [as] to any material fact and the movant is entitled to judgment as a matter of law." U.S. ex. rel. Jamison v. McKesson Corp., 649 F.3d 322, 326 (5th Cir.2011).

3. IMA's Interpretation of the Plans

We limit our review of the interpretation of a benefits plan under ERISA to the administrative record. See Vega v. Nat'l Life Ins. Servs., Inc., 188 F.3d 287, 299 (5th Cir.1999) (en banc), overruled on other grounds by Metro. Life Ins. Co. v. Glenn, 554 U.S. 105, 128 S.Ct. 2343, 171 L.Ed.2d 299 (2008); Estate of Bratton v. Nat'l Union Fire Ins. Co. of Pittsburgh, 215 F.3d 516, 521 (5th Cir.2000). In evaluating the record to determine whether the interpretation of a plan is "legally correct," we consider: "(1) whether the administrator has given the plan a uniform construction, (2) whether the interpretation is consistent with a fair reading of the plan, and (3) any unanticipated costs resulting from different interpretations of the plan." Crowell v. Shell Oil Co., 541 F.3d 295, 312 (5th Cir.2008). "[W]hether the administrator gave the plan a fair reading is the most important factor." Stone v. UNOCAL Termination Allowance Plan, 570 F.3d 252, 260 (5th Cir.2009); see also Crowell, 541 F.3d at 313. An administrator's interpretation is consistent with a fair reading of the plan if it construes the plan according to the "plain meaning of the plan language." Threadgill v. Prudential Sec. Grp., Inc., 145 F.3d 286, 292 (5th Cir.1998); see also Stone, 570 F.3d at 260.

If this court finds that an administrator's interpretation of a plan is incorrect, then we consider whether the interpretation was an abuse of discretion. Chacko v. Sabre, Inc., 473 F.3d 604, 611 (5th Cir.2006); see also Crowell, 541 F.3d at 312. A plan administrator abuses its discretion "[w]ithout some concrete evidence in the administrative record that supports the denial of the claim." Vega, 188 F.3d at 302. Abuse of discretion factors include: "(1) the internal consistency of the plan under the administrator's interpretation, (2) any relevant regulations formulated by the appropriate administrative agencies, and (3) the factual background of the determination and any inferences of lack of good faith." Gosselink v. Am. Tel. & Tel., Inc., 272 F.3d 722, 726 (5th Cir. 2001). However, "if an administrator interprets an ERISA plan in a manner that directly contradicts the plain meaning of the plan language, the administrator has abused his discretion even if there is neither evidence of bad faith nor of a violation of any relevant administrative regulations." Id. at 727.

Here, IMA's interpretation of the plans was incorrect because its finding that LifeCare was an SNF was inconsistent with a fair reading of the plans.4 The plans state that an SNF "is a facility that fully meets all of" the seven SNF tests. IMA has acknowledged that, absent its interpretation of the final sentence, a facility must meet all seven factors to qualify as an SNF. Yet IMA's denial of Evans' full claim references only two of the seven factors; its denial of Wall's claim does not reference a single factor.

IMA instead argues that it interpreted the plans correctly by categorizing LifeCare as an SNF under what IMA contends is an alternative definition of an SNF, which IMA contends is independent of the seven factors, which an SNF otherwise would "fully" have to meet. IMA contends that this final sentence in fact is an "alternative" and "independent" second definition of an SNF that allows it to categorize a medical provider as an SNF if the provider refers to itself as an LTAC or rehab hospital, and that LifeCare referred to itself as both. IMA observes that it denied Wall's claim by stating that LifeCare was "a rehab facility as defined in the plan." IMA adds that it denied Evans' full claim by quoting the final sentence, and by describing LifeCare as "an extended care/long term acute care facility."

However, a fair reading of these specific plans shows that the final sentence does not permit IMA to categorize LifeCare as an SNF solely because LifeCare refers to itself as an LTAC. First, the plain language of the sentence — "[t]his [SNF] term also applies to charges incurred in a facility referring to itself as" an LTAC — provides that a facility cannot be excepted from classification as an SNF merely by referring to itself as an LTAC instead of an SNF. The sentence, by its plain language and logically, clarifies that an SNF by any other name is still an SNF: a facility must "fully meet[] all of" the seven factors to qualify as an SNF, even if it refers to itself as an LTAC. Notably, the sentence in question, by its explicit language, clarifies that the "term" SNF encompasses facilities that use nomenclature other than SNF. The final, clarifying sentence, with its antecedent being the "term" SNF, offers no "alternative" or "independent" or "second" catch-all definition of an SNF.

To the extent that the plans' texts were ambiguous, Bennett's testimony supports our reading of its terms.5 Bennett testified that she denied Evans' and Wall's claims because LifeCare qualified as an SNF under the plans' seven-factor test. IMA challenges this reading of Bennett's deposition testimony as "tortured." IMA instead argues that Bennett said only that LifeCare was an SNF under the final sentence and that, alternatively, she did not understand counsel's question. However, the transcript of the deposition shows that counsel's questions and Bennett's answers were clear.6 Otherwise, a facility that happened to describe itself as a "Long-Term Acute Care" provider would qualify as an SNF under the final sentence regardless of the type of services it provided.

A fair reading of the plans shows that IMA's interpretation categorizing LifeCare as an SNF without applying each of the seven SNF factors was incorrect, as found by the district court. As a result, we must address whether this incorrect interpretation is an abuse of discretion.

IMA's interpretation of the plans was an abuse of discretion because IMA's categorization of LifeCare as an SNF "directly contradict[ed] the plain meaning of the plan language" under the "factual background" abuse of discretion prong. See Gosselink, 272 F.3d at 726-27. As discussed above, the plain language of the plans provides that, even if a medical provider refers to itself as an LTAC, it may be an SNF entitled to limited or no reimbursement if it "fully meets all of" the seven factors set forth. IMA's interpretation of the plans to categorize LifeCare as an LTAC solely because it referred to itself as one contradicted the plans' plain language. Accordingly, we do not need to consider the other two abuse of discretion factors. See Gosselink, 272 F.3d at 727. LifeCare did not provide "concrete evidence" that LifeCare met each of the seven factors. IMA's letter denying Evans' claim referenced only two of the seven factors; its letter denying Wall's claim did not mention a single factor. IMA's counsel forthrightly acknowledged: "... I don't have anything in the record that says we have gone through the seven-element analysis...."

In sum, IMA incorrectly interpreted the plans because it categorized LifeCare as an SNF without finding that LifeCare "fully meets all of" the plans' seven SNF factors. IMA abused its discretion because categorizing LifeCare as an SNF without considering the seven-factor SNF test contradicted the plain language of the plans.

4. IMA's Liability

An ERISA claimant may bring a lawsuit under 29 U.S.C. § 1132(a)(1)(B) "to recover benefits due to him under the terms of his plan." Schadler v. Anthem Life Ins. Co., 147 F.3d 388, 394 (5th Cir. 1998). This court has found that a claimant may bring a suit against an employer when the plan has no meaningful existence apart from the employer, and when the employer made the decision to deny benefits. Musmeci v. Schwegmann Giant Super Markets, Inc., 332 F.3d 339, 349-50 (5th Cir.2003).7

We start with the language of the statute. The plain language of § 1132(a)(1)(B) — permitting an action "to recover benefits due to him under the terms of his plan, to enforce his rights under the terms of the plan, or to clarify his rights to future benefits under the terms of the plan" — does not limit the scope of defendants that a claimant may bring a lawsuit against.8

We next look to our sister circuits. At least four circuits have found that entities other than the benefits plan or the employer plan administrators may be held liable under § 1132(a)(1)(B). See Cyr v. Reliance Standard Life Ins. Co., 642 F.3d 1202, 1206 (9th Cir.2011) (en banc) (finding that "potential defendants in actions brought under § 1132(a)(1)(B) should not be limited to plans and plan administrators"); Mein v. Carus Corp., 241 F.3d 581, 585 (7th Cir.2001) ("While it is silly not to name the plan as a defendant in an ERISA suit, we see no ... reason to have this case stand starkly for the proposition that the plan is always the only proper defendant...."); Layes v. Mead Corp., 132 F.3d 1246, 1249-50 (8th Cir.1998) (finding that a non-employer plan administrator with discretionary authority could be held liable); Daniel v. Eaton Corp., 839 F.2d 263, 266 (6th Cir.1988) (same).

The Appellants argue that the Seventh and Eleventh Circuits have limited liability under § 1132(a)(1)(B), but the cases they rely on are distinguishable from the facts here. In Howard v. Parisian, Inc., 807 F.2d 1560, 1564-65 (11th Cir.1987), the Eleventh Circuit addressed only whether ERISA preempts state law claims against non-fiduciary plan administrators, not whether a TPA may be held liable under § 1132(a)(1)(B). In Baker v. Big Star Division of the Grand Union Co., 893 F.2d 288, 290 (11th Cir.1989), the administrator did not have the authority to deny benefits. Likewise, in Harris Trust & Savings Bank v. Provident Life & Accident Insurance Co., 57 F.3d 608, 613 (7th Cir.1995), the administrator did not exercise discretion over the claims process.9

Notably, courts finding liability under § 1132(a)(1)(B) nonetheless apply a restrained functional test: a party will be exposed to liability only if it exercises "actual control" over the administration of the plan. See Musmeci, 332 F.3d at 349-50 (finding that an employer that makes benefits decisions, and has no meaningful existence apart from the plan, may be held liable); Garren v. John Hancock Mut. Life Ins. Co., 114 F.3d 186, 187 (11th Cir.1997) (per curiam) ("The proper party defendant in an action concerning ERISA benefits is the party that controls administration of the plan."); Gore v. El Paso Energy Corp. Long Term Disability Plan, 477 F.3d 833, 842 (6th Cir.2007) (finding that an employer administrator was not liable because it did not control the claims process). Courts extending liability to entities other than a benefits plan or an employer plan administrator likewise apply this functional approach. Layes, 132 F.3d at 1249-50 (finding that an administrator with discretionary authority could be held liable); Daniel, 839 F.2d at 266 (same); Pippin v. Broadspire Servs., Inc., No. Civ.A. 05-2125, 2006 WL 2588009, at *2 (W.D.La. Sept. 8, 2006) (finding that "an examination of [the TPA's] role in denying [the plaintiff's] benefits claim is essential in order to determine whether it is a proper party"); Kellebrew v. UNUM Life Ins. Co. of Am., Civ. Action No. H-06-0275, 2006 WL 1050664, at *2 (S.D.Tex. Apr. 20, 2006) (refusing to dismiss a benefits claim against an administrator that "actually administers" the plan).

We find the rationale and cases holding that a TPA may be held liable only if it exercises "actual control" over the benefits claims process convincing. We agree that "[t]he proper party defendant in an action concerning ERISA benefits is the party that controls administration of the plan" and that "[i]f an entity or person other than the named plan administrator takes on the responsibilities of the administrator, that entity may also be liable for benefits." Gomez-Gonzalez v. Rural Opportunities, Inc., 626 F.3d 654, 665 (1st Cir.2010) (internal citations omitted). Neither the statute nor caselaw directs that § 1132(a)(1)(B) should insulate an entity from liability merely for being a TPA. See Harris Trust, 530 U.S. at 245, 120 S.Ct. 2180 (rejecting argument that "absent a substantive provision of ERISA expressly imposing a duty upon a nonfiduciary party in interest, the nonfiduciary party may not be held liable under ... one of ERISA's remedial provisions"). Where a TPA exercises control over a plan's benefits claims process, and exerts that control to deny a claim by incorrectly interpreting a plan in a way that amounts to an abuse of discretion, liability may attach. See Cyr, 642 F.3d at 1207 (extending liability under § 1132(a)(1)(B) to entity responsible for paying legitimate claims where plan administrator "had nothing to do with denying [plaintiff's] claim for increased benefits").

As a result, we proceed to consider whether IMA exercised actual control over the denial of Evans' and Wall's claims. Here, the administration contracts between Carter and BRI provided that:

"[T]he mere exercise of physical control or the performance of mechanical administrative tasks generally is insufficient" for liability under § 1132(a)(1)(B). Gomez-Gonzalez, 626 F.3d at 665. However, IMA also had authority to "[p]rocess all claims presented for benefit under [the] Plan." IMA acknowledged that it would not consult with BRI or Carter to resolve a claim unless a "gray area" presented itself. IMA also admitted that it determined that Evans' and Wall's claims were "routine" and therefore did not refer them to BRI or Carter. IMA was thus responsible for, first, interpreting the plans to determine whether the claims at issue were routine or non-routine, and, second, interpreting the terms of the plans to deny Evans and Wall benefits. See IT Corp. v. Gen. Am. Life Ins. Co., 107 F.3d 1415 (9th Cir.1997) ("[I]t is hard to say that [the claims processor] has no power to make decisions about plan interpretation, because [the claims processor] has to interpret the plan to determine whether a benefits claim ought to be referred back."). In so doing, IMA's actions distinguish it from those cases in which administrators were found not liable for performing only non-discretionary functions. See Provident Life & Acc. Ins. Co., 57 F.3d at 613 (finding claims administrator not liable where administrator could "elect to advance benefits" but it was employer who "retain[ed] the right ... to decide all disputed and non-routine claims"); Baker, 893 F.2d at 290 (claims processor not liable where it "ha[d] not been granted the authority to review benefits denials and make the ultimate decisions regarding eligibility").

This case would be different had the administration contracts not given IMA the power to deny claims IMA considered routine. Had IMA referred all disputed claims to BRI and Carter for resolution it would not now be liable for having exercised discretionary authority in denying Evans' and Wall's benefits claims. Alternatively, if the administrative record had included evidence that BRI and Carter had furnished IMA with an interpretation of the term "skilled nursing facility," IMA might credibly have argued that it did not apply its own interpretation, but only applied that of the plan administrators.

We find that the district court correctly held that LifeCare could maintain an action against IMA pursuant to § 1132(a)(1)(B) and that IMA was liable for exercising actual control over the claims process.10

5. LifeCare's Attorneys' Fees Award

This court reviews an award of attorneys' fees for abuse of discretion, reviewing factual findings for clear error and legal conclusions de novo. Dearmore v. City of Garland, 519 F.3d 517, 520 (5th Cir.2008).

Pursuant to 29 U.S.C. § 1132(g)(1) of ERISA, this court "in its discretion may allow a reasonable attorney's fee and costs of action to either party" so long as the party has achieved "some degree of success on the merits." Hardt v. Reliance Standard Life Ins. Co., ___ U.S. ___, 130 S.Ct. 2149, 2151, 176 L.Ed.2d 998 (2010) ("This Court's `prevailing party' precedents do not govern here because that term of art does not appear in § 1132(g)(1)."). A party satisfies this "success on the merits" requirement "if the court can fairly call the outcome of the litigation some success on the merits without conducting a lengthy inquir[y] into the question whether a particular party's success was `substantial' or occurred on a `central issue.'" Id. at 2158 (internal quotations omitted).

This court has assessed attorney's fees under ERISA in the past by applying the five-factor test from Iron Workers Local No. 272 v. Bowen, 624 F.2d 1255, 1266 (5th Cir.1980). However, the Supreme Court has clarified that we do not need to consider the Bowen factors. Hardt, 130 S.Ct. at 2158 ("Because these five factors bear no obvious relation to § 1132(g)(1)'s text or to our fee-shifting jurisprudence, they are not required for channeling a court's discretion when awarding fees under this section."); see also 1 Lincoln Fin. Co. v. Metro. Life Ins. Co., 428 Fed.Appx. 394, 396 (5th Cir. 2011) (per curiam) (unpublished) ("A district court may consider the five factors, but Hardt does not mandate consideration.").

Here, the Appellants advance several arguments against the district court's award of attorneys' fees to LifeCare. The Appellants observe that the district court did not discuss the Bowen factors. As the Supreme Court made clear in Hardt, however, the Bowen factors are discretionary. See 130 S.Ct. at 2158. More fundamentally, the Appellants argue that the district court erred by awarding excessive attorneys' fees. Specifically, Appellants contend that this court should reduce LifeCare's $453,000 attorneys' fees award by $150,000 because the district court awarded $80,000 in fees for work on dismissed state law claims, $50,000 in fees for the allegedly generic time entries, and $20,000 in fees for pre-suit attorneys' fees.

The award of $80,000 in fees for work on dismissed state law claims was not an abuse of discretion because LifeCare achieved "some degree of success on the merits" in the overall litigation.11 Hardt, 130 S.Ct. at 2151. Further, after a careful review of the record, the district judge reduced the requested award by $30,000 for LifeCare's work on the state law claims. The award of $50,000 in fees for generic time entries was not an abuse of discretion because LifeCare provided extensive billing entries that included a description of each entry, and the time spent on each task. Likewise, the award of $20,000 in fees for pre-suit work was not an abuse of discretion because the Appellants have failed to show that the fees were not for work in preparation for this lawsuit.12

Additionally, the Appellants argue that the district court erred by awarding $65,000 in conditional appellate attorneys' fees to LifeCare because there was "no evidence submitted" to support such fees.13 However, the award of conditional fees was not an abuse of discretion because LifeCare produced evidence in the form of a detailed affidavit by LifeCare's counsel explaining why the fees were necessary.14

6. Conclusion

Accordingly, we AFFIRM the district court's judgment and award of attorneys' fees.