KING, Circuit Judge:
In this qui tam action under the False Claims Act (the “FCA”), defendant Agape Senior Community, Inc., and the twenty-three other defendants (collectively, “Agape”) are affiliated entities that operate elder care facilities throughout South Carolina.
The relators, Brianna Michaels and Amy Whitesides, are former Agape employees who allege that Agape fraudulently billed Medicare and other federal health care programs for services to thousands of patients — services that were not actually provided, or that were provided to patients who were not eligible for them. The United States Government was entitled, but declined, to intervene.
To establish liability and damages, the relators sought to rely on statistical sampling. The district court determined, however, that using statistical sampling to prove their case would be improper (the “statistical sampling ruling”). Additionally, the court rejected a proposed settlement between the relators and Agape, because the Attorney General of the United States objected to it. In so doing, the court concluded that the Government — despite not having intervened in an FCA qui tam action — possesses an unreviewable veto authority over the action’s proposed settlement (the “unreviewable veto ruling”).
The district court certified both its statistical sampling and unreviewable veto rulings for these interlocutory appeals under 28 U.S.C. § 1292(b). We thereafter granted the petitions for permission to appeal submitted to this Court by the rela-tors (seeking an appeal from both rulings) and by Agape (requesting an appeal from the unreviewable veto ruling only). As explained below, we affirm the unreviewable veto ruling and dismiss as improvidently granted the relators’ appeal as to the statistical sampling ruling.
I.
A.
The FCA, codified at 31 U.S.C. §§ 3729-3733, authorizes a private individual (i.e., a relator) to initiate and pursue an action in the name of the United States Government (a qui tam action) to seek civil remedies for fraud against the Government.
See
31 U.S.C. § 3730(b)(1). Pursuant to § 3730(b)(1), the qui tam “action may be dismissed only if the court and the Attorney General give written consent to the dismissal and their reasons for consenting.”
At the outset of the qui tam action, the relator’s complaint must be served on the Government, filed
in camera,
and kept under seal for at least sixty days, with no service of process on the defendant until the court so orders.
See
31 U.S.C. § 3730(b)(2). During the sixty-day period after it receives the complaint, the Government may elect to intervene in the qui tam action.
Id.
Specifically, before the expiration of the sixty-day period — or any exten
sion thereof under § 3730(b)(3) — the Government must either (A) “proceed with the action” by assuming primary responsibility for the action’s prosecution, or (B) “notify the court that it declines to take over the action” from the relator, who will then “have the right to conduct the action.”
Id.
§ 3730(b)(4)(A)-(B). If the Government declines to intervene during the initial sixty-day (or extended) period, the court may nevertheless permit its intervention “at a later date upon a showing of good cause.”
Id.
§ 3730(c)(3).
Once the Government intervenes, the relator retains the right to continue as a party to the action, subject to certain limitations.
See
31 U.S.C. § 3730(c)(1). For example, the Government is authorized to settle the action over the relator’s objection, but only “if the court determines, after a hearing, that the proposed settlement is fair, adequate, and reasonable under all the circumstances.”
Id.
§ 3730(c)(2)(B).
When the qui tam action is successful, the relator is entitled to share with the Government in the award.
See
31 U.S.C. § 3730(d)(l)-(4). The amount of. the relator’s share depends on whether the Government intervened in the action. If the Government did not intervene, “the person bringing the action or settling the claim shall receive an amount which the court decides is reasonable for collecting the civil penalty and damages.”
Id.
§ 3730(d)(2) (specifying that such “amount shall be not less than 25 percent and not more than 30 percent of the proceeds of the action or settlement”).
B.
Here, the relators served their initial Complaint on the Government and, on December 7, 2012, filed it under seal in the District of South Carolina. The district court extended the Government’s deadline for its intervention decision to March 5, 2013. By its notice of that date, the Government declined to intervene but called attention to the consent-for-dismissal provision of § 3730(b)(1), requesting that the relators and Agape solicit the Attorney General’s written consent before asking the court to rule on any proposed dismissal. Two days later, on March 7, 2013, the court unsealed the Complaint and directed the relators to serve it on Agape.
The relators filed their operative Second Aménded Complaint on March 6, 2014, and discovery ensued.
Although the relators and Agape dispute the exact numbers, they agree that Agape admitted more than 10,000 patients to its facilities in South Carolina and submitted more than 50,000 claims to federal health care programs during the relevant time period. The rela-tors sought to use statistical sampling to prove their case in order to avoid the cost of reviewing each patient’s chart to identify which claims were fraudulent — a task that the relators said would take their experts four to nine hours per patient, at a rate of $400 per hour, potentially totalling more than $36 million. For its part, Agape opposed the use of any evidentiary form of statistical sampling. Thus, the district court received briefing and conducted a hearing on the issue. By Order of March 16, 2015, the court made its statistical sampling ruling “that based on the facts of this case, statistical sampling would be improper.”
See United States ex rel. Michaels v. Agape Senior Cmty., Inc.,
No. 0:12-cv-03466, at 2 (D.S.C. Mar. 16, 2015), ECF No. 255 (the “March 2015 Order”).
Meanwhile, the relators, Agape, and the Government had mediated unsuccessfully in November 2014, and the relators and Agape had mediated again — without the Government’s knowledge — in January 2015. The proposed settlement between the relators and Agape emerged from the second mediation. Relying on § 3780(b)(1), the Attorney General objected to the proposed settlement. The Government has not, however, sought permission pursuant to § 3730(c)(3) to intervene in this action, -thus standing by its initial decision under § 3730(b)(2)-(4) to decline intervention.
In objecting to the proposed settlement, the Attorney General protested in part that the settlement amount was appreciably less than $25 million, the Government’s estimate of total damages based on its own use of statistical sampling. The various bases for the Attorney General’s objection were stated and discussed during a series of status conferences conducted by the district court in an effort to determine if this action could be settled.
Agape eventually filed a motion to enforce the proposed settlement over the Attorney General’s objection, and the Government filed a response opposing that motion. By Order of June 25, 2015, the district court rendered its unreviewable veto ruling and thereby sustained the Attorney General’s objection to the proposed settlement.
See United States ex rel. Michaels v. Agape Senior Cmty., Inc.,
No. 0:12-cv-03466, 2015 WL 3903675 (D.S.C. June 25, 2015), EOF No. 296 (the “June 2015 Order”). The June 2015 Order also expounded on the statistical sampling ruling that had been made in the March 2015 Order. Additionally, the June 2015 Order certified
sua sponte
both the unreviewable veto and statistical sampling rulings for these interlocutory appeals under 28 U.S.C. § 1292(b).
The district court prefaced its analysis in the June 2015 Order with a description of the “unique dilemma” that it faced:
The Government, claiming an unreviewable veto right over the tentative settlement in this case, objects to a settlement in a case to which it is not a party, using as a basis of its objection some form of statistical sampling that this Court has rejected for use at the trial of the case.
See
June 2015 Order 6. In rendering its unreviewable veto ruling, the district court rejected the argument supporting the proposed settlement that the relators and Agape jointly advanced in reliance on the Ninth Circuit’s decision in
United States ex rel. Killingsworth v. Northrop Corp.,
25 F.3d 715 (9th Cir. 1994). Their argument was that, because the Government had declined to intervene herein, the Attorney General’s objection to the proposed settlement was subject to the district court’s reasonableness review. The district court instead agreed with the Government — as well as the Fifth Circuit in
Searcy v. Philips Electronics North America Corp.,
117 F.3d 154 (5th Cir. 1997), and the Sixth Circuit in
United States v. Health Possibilities, P.S.C.,
207 F.3d 335 (6th Cir. 2000)—that the Attorney General possesses an absolute veto power over voluntary settlements in FCA qui tarn actions. In so ruling, the district court explained that it was adhering to the plain language of 31 U.S.C. § 3730(b)(1), which “provides no limitation on the Attorney General’s authority, and no right of [a court] to review the Attorney General’s objection for reasonableness.”
See
June 2015 Order 6-7.
Nevertheless, the district court noted that, if it “did have the authority to review an objection by the Attorney General for reasonableness in a case of this nature, a compelling case could be made here that the Government’s position is not, in fact, reasonable.”
See
June 2015 Order 10. Such a compelling case would include that the relators “could be looking at an expenditure of between $16.2 million and $36.5 million in pretrial preparation alone for a case that the Government values at $25 million.”
Id.
at 11. Moreover, as the court observed, although “the Government has admitted that statistical sampling of the entire universe of claims played a major part in its calculation of the value of this case,” it resisted (with the court’s reluctant approval) discovery requests seeking specific details about its calculation.
Id.
at 12.
Turning to its earlier statistical sampling ruling, the district court spelled out its rationale for concluding that it would be improper to use statistical sampling evidence to prove the relators’ case. In sum, the court explained that statistical sampling can be appropriate “where the evidence has dissipated, thus rendering direct proof of damages impossible.”
See
June 2015 Order 13-14 (citing example of FCA qui tam action where defendant allegedly defrauded Government in moving household belongings of military personnel by artificially bumping weight of shipments that had since been completed). Here, however, “nothing has been destroyed or dissipated.... The patients’ medical charts are all intact and available for review by either party.”
Id.
at 14.
Finally, in certifying its unreviewable veto and statistical sampling rulings for these interlocutory appeals, the district court observed that the relators and Agape “face a trial of monumental proportions, involving a staggering outlay of expenses by the [relators] and a significant drain of [court] resources.”
See
June 2015 Order 18. The court deemed it to “be much more judicially efficient to have a ruling on both of the questions before, rather than after, such a monumental trial.”
Id.
Echoing the requirements of 28 U.S.C. § 1292(b), the court also stated that each ruling involves “a controlling question of law as to which there is substantial ground for difference of opinion, and that an immediate appeal ... may materially advance the ultimate termination of this litigation.”
Id.
at 19. Because we granted the relators’ and Agape’s subsequent petitions for permission to appeal, we possess jurisdiction pursuant to § 1292(b).
II.
We first assess the district court’s unreviewable veto ruling. Because the interpretation of 31 U.S.C. § 3730 presents a pure question of law, our review is de novo.
See United States ex rel. Oberg v. Pa. Higher Educ. Assistance Agency,
804 F.3d 646, 654 (4th Cir. 2015). Our focus, of course, is on § 3730(b)(1), which provides in full:
A person may bring a civil action for a violation of [the FCA] for the person and for the United States Government. The action shall be brought in the name of the Government.
The action may be dismissed only if the court and the Attorney General give written consent to
the dismissal and their reasons for consenting.
31 U.S.C. § 3730(b)(1) (emphasis added).
1..
The question presented — the extent of the Attorney General’s power under § 3730(b)(1) to veto the voluntary settlement of an FCA qui tam action in which the Government declined to intervene — is not one that we have heretofore squarely confronted.
Thus, it is helpful to begin with a discussion of the three courts of appeals decisions debated in the district court:
United States ex rel. Killingsworth v. Northrop Corp.,
25 F.3d 715 (9th Cir. 1994);
Searcy v. Philips Electronics North America Corp.,
117 F.3d 154 (5th Cir. 1997); and
United States v. Health Possibilities, P.S.C.,
207 F.3d 335 (6th Cir. 2000).
In the
Killingsworth
decision, the Ninth Circuit determined that § 3730(b)(l)’s consent-for-dismissal provision is limited by § 3730(b)(2) — (4), which delineates the initial sixty-day (or extended) period during which the Government may elect to intervene, as well as by § 3730(c)(3), which authorizes the court to permit later intervention upon a showing of good cause.
See
25 F.3d at 722. The
Killingsworth
court ruled that “the consent provision contained in § 3730(b)(1) applies only during the initial sixty-day (or extended) period.”
Id.
Thereafter, the Government’s settlement-related authority depends on whether it has intervened, i.e., whether the Government or the relator is empowered to control the litigation.
Id.
When the Government has not intervened,
Killingsworth
merely permits the Attorney General to object with “good cause” to a proposed settlement and obtain a hearing on whether the settlement is “fair and reasonable.”
Id.
at 723-25 (cobbling standard from § 3730(c)(2)(B), § 3730(c)(3), and other aspects of § 3730).
The Ninth Circuit resolved in Killings-
worth
that the Government’s position— “that without intervention [the Attorney General] possesses an absolute right to reject a proposed settlement at any time and for any reason” — cannot comport with the plain language of § 3730(b)(4)(B), which affords the relator “the right to conduct the action” once the Government has declined to intervene.
See
25 F.3d at 722. According to the court, that is because “[t]he right to conduct a qui tam action obviously includes the right to negotiate a settlement in that action.”
Id.
For that proposition, the court relied on § 3730(d)(2), which .provides that if the Government has not intervened, “the person bringing the action or settling the claim shall receive an amount which the court decides is reasonable for collecting the civil penalty and damages.” The
Kill-ingsworth
decision emphasized the “or settling the claim” language of § 3730(d)(2), propounding that it “confirms the relator’s right to settle the action if the government declines to intervene.”
See
25 F.3d at 722-23.
The Ninth Circuit’s interpretation in
Killingsworth
of § 3730 was subsequently rejected by the Fifth Circuit in its
Searcy
decision and the Sixth Circuit in its
Health Possibilities
decision. Unlike the Ninth Circuit, those latter two courts recognized “an absolute veto power over voluntary settlements in qui tam [FCA] suits,”
see Searcy,
117 F.3d at 158, under which a relator “may not seek a voluntary dismissal of any action ... without the Attorney General’s consent,”
see Health Possibilities,
207 F.3d at 336.
As the Fifth Circuit explained in
Searcy,
the language of § 3730(b)(l)’s consent-for-dismissal provision “is as unambiguous as one can expect,” and there is “nothing in § 3730 to negate [that language’s] plain import.”
See
117 F.3d at 159. In particular, the
Searcy
court confronted those aspects of § 3730 utilized in
Killingsworth:
§ 3730(b)(4)(B) (according the relator “the right to conduct the action” when the Government declines to intervene), and § 3730(d)(2) (providing a reasonable amount to “the person bringing the action or settling the claim”). More specifically,
Searcy
refuted
Killingsworth’&
pronouncements that the § 3730(b)(4)(B) “right to conduct a qui tam action obviously includes the right to negotiate a settlement in that action,” and that § 3730(d)(2) “confirms the relator’s right to settle the action if the government declines to intervene.”
See Killingsworth,
25 F.3d at 722-23. The
Se-arcy
court expounded that “[a] relator has ‘conducted’ an action if he devises strategy, executes discovery, and argues the case in court, even if the government frustrates his settlement efforts.”
See Searcy,
117 F.3d at 160. Moreover, the court observed that “the government’s power to block settlements does not mean that the relator will never be the person settling the claim.”
Id.
The
Searcy
court further recognized “that relators can manipulate settlements in ways that unfairly enrich them and reduce benefits to the government,” including “by bargaining away claims on behalf of the United States.”
See
117 F.3d at 160. Section 3730(b)(l)’s consent-for-dismissal provision, however, “allows the government to resist [such] tactics and protect its ability to prosecute matters in the future.”
Id.
Along those same lines, the Sixth Circuit observed in
Health Possibilities
that “the power to veto a privately negotiated settlement of public claims is a critical aspect of the government’s ability to protect the public interest in qui tam litigation. The FCA is not designed to serve the parochial interests of relators, but to vindicate civic interests in avoiding fraud against public monies.”
See
207 F.3d at 340. The
Health Possibilities
court underscored that “[t]he location of the consent provision [in § 3730(b)(1) ] immediately after the command that the action be brought in the government’s name suggests that it is an important component of the government’s ability to regulate qui tam actions.”
Id.
at 342.
Notably, the Fifth, Sixth, and Ninth Circuits considered the legislative history of the FCA. On the one hand, the Ninth Circuit discerned a congressional “intent to place full responsibility for [FCA] litigation on private parties, absent early intervention by the government or later intervention for good cause” — an intent that the court deemed to be “fundamentally inconsistent with the asserted ‘absolute’ right of the government to block a settlement and force a private party to continue litigation.”
See Killingsworth,
25 F.3d at 722.
On the other hand, the Fifth and Sixth Circuits perceived Congress’s intent to grant the Attorney General full veto authority that has existed since the original FCA statute was enacted in 1863 during the Civil War.
See Health Possibilities,
207 F.3d at 342-43;
Searcy,
117 F.3d at 159. As those courts saw it, that intent has endured even through subsequent amend
ments to the FCA providing more incentives to relators and creating and expanding the Government’s power to intervene.
Id.
Those courts thus concluded:
For more than 130 years, Congress has instructed courts to let the government stand on the sidelines and veto a voluntary settlement. It would take a serious conflict within the structure of the [FCA] or a profound gap in the reasonableness of the [consent-for-dismissal] provision for us to be able to justify ignoring this language. We can find neither.
Health Possibilities,
207 F.3d at 344 (quoting
Searcy,
117 F.3d at 160). Here, in rendering its unreviewable veto ruling, the district court similarly interpreted § 3730(b)(1) and its consent-for-dismissal provision.
2.
We agree with the district court, and with the Fifth and Sixth Circuits, that the Attorney General possesses an absolute veto power over voluntary settlements in FCA qui tam actions. In reaching that conclusion, we rely on the plain language of 31 U.S.C. § 3730(b)(1), “readpng] the words in their context and with a view to their place in the overall statutory scheme.”
See King v. Burwell,
— U.S. -, 135 S.Ct. 2480, 2489, 192 L.Ed.2d 483 (2015) (internal quotation marks omitted). Simply put, nothing else in § 3730 leads us to doubt that Congress meant exactly what it said in § 3730(b)(1) — that a qui tam action “may be dismissed only if the court and the Attorney General give written consent to the dismissal and their reasons for consenting.”
On appeal, neither the relators nor Agape advocate the Ninth Circuit’s theory that the consent-for-dismissal provision “applies only during the initial sixty-day (or extended) period” in which the Government must decide whether to intervene in a qui tam action.
See Killingsworth,
25 F.3d at 722. Somewhat like the Ninth Circuit, however, Agape contends that the Government cannot unreasonably withhold its consent to a settlement. According to Agape, the reasonableness requirement flows from § 3730(b)(1) itself.
See
Br. of Agape 20 (“[A] correct reading of § 3730(b)(1) recognizes that the Government’s consent to a qui tam settlement cannot be unreasonably withheld.”). For its part, the Ninth Circuit cobbled a standard from other aspects of § 3730, including § 3730(c)(2)(B) and § 3730(c)(3), limiting the Attorney General to an objection for “good cause” and a hearing on whether the proposed settlement is 'f‘fair and reasonable.”
See Killingsworth,
25 F.3d at 723-25. Both Agape and the Ninth Circuit have reasoned that an unlimited veto power cannot coexist with § 3730(b)(4)(B) insofar as it confers on the relator “the right to conduct the action” when the Government declines to intervene, or with § 3730(d)(2) insofar as it provides for a share of the award to “the person bringing the action or settling the claim.”
Of course, as the Fifth Circuit deftly explained, the right to conduct.the action does not necessarily include the right to settle the claim, although, absent the Attorney General’s objection, the relator may yet settle the claim.
See Searcy,
117 F.3d at 160. That is, § 3730(b)(4)(B) and § 3730(d)(2) cannot reasonably be understood to create an unfettered right to settle on the part of the relator.
Furthermore, § 3730(b)(1) and its consent-for-dismissal provision is not temporally qualified or explicitly limited in any other manner. Unlike other provisions of § 3730, § 3730(b)(1) does not overtly require the Government to satisfy any standard or make any showing reviewable by the court. A prime example is
§ 3730(c)(2)(B), under which the Government may settle a qui tam action over the relator’s objection, but only “if the court determines, after a hearing, that the proposed settlement is fair, adequate, and reasonable under all the circumstances.” Congress could have readily included similar language in § 3730(b)(1); that it decided against doing so is enlightening.
See Barnhart v. Sigmon Coal Co.,
534 U.S. 438, 452, 122 S.Ct. 941, 151 L.Ed.2d 908 (2002) (“[Wjhen Congress includes particular language in one section of a statute but omits it in another section of the same Act, it is generally presumed that Congress acts intentionally and purposely in the disparate inclusion or exclusion.” (internal quotation marks omitted)).
Finally, we would be remiss not to recognize that the Attorney General’s absolute veto authority is entirely consistent with the statutory scheme of the FCA. Even where the Government declines to intervene, “the United States is the real party in interest in any [FCA] suit.”
See United States ex rel. Milam v. Univ. of Tex. M.D. Anderson Cancer Ctr.,
961 F.2d 46, 50 (4th Cir. 1992). Meanwhile, “[a]s a class of plaintiffs, qui tam relators are different in kind than the Government. They are motivated primarily by prospects of monetary reward rather than the public good.”
See Hughes Aircraft Co. v. United States ex rel. Schumer,
520 U.S. 939, 949, 117 S.Ct. 1871, 138 L.Ed.2d 135 (1997). Instead of freeing relators to maximize their own rewards at the public’s expense, Congress has granted the Attorney General the broad and unqualified right to veto proposed settlements of qui tam actions.
Accordingly, we reject Agape’s interpretation of § 3730 and conclude today that, under the plain language of § 3730(b)(1), the Attorney General possesses an absolute veto power over voluntary settlements in FCA qui tam actions. The district court having concluded the same, we affirm its unreviewable veto ruling.
Turning to the district court’s statistical sampling ruling, we find it prudent to re-examine whether that aspect of the relator’s appeal is appropriate for interlocutory review under 28 U.S.C. § 1292(b). Pursuant thereto, the order being reviewed must involve “a controlling question of law as to which there is substantial ground for difference of opinion,” and an immediate appeal from that order must promise to “materially advance the ultimate termination of the litigation.” We have cautioned “that § 1292(b) should be used sparingly and thus that its requirements must be strictly construed.”
See Myles v. Laffitte,
881 F.2d 125, 127 (4th Cir. 1989).
Strictly construing § 1292(b), we recognize that it may be proper to conduct an interlocutory review of an order presenting “a pure question of law,” i.e., “an abstract legal issue that the court of appeals can decide quickly and cleanly.”
See Mamani v. Berzain,
825 F.3d 1304, 1312 (11th Cir. 2016) (internal quotation marks omitted). In other words, § 1292(b) review may
be appropriate where “the court of appeals can rule on a pure, controlling question of law without having to delve beyond the surface of the record in order to determine the facts.”
See McFarlin v. Conseco Servs., LLC,
381 F.3d 1251, 1259 (11th Cir. 2004). Such a pure question of law includes the issue raised in the relators’ and Agape’s appeals from the district court’s unreviewable veto ruling.
By contrast, § 1292(b) review is not appropriate where, for example, the question presented “turns on whether there is a genuine issue of fact or whether the district court properly applied settled law to the facts or evidence of a particular case.”
See McFarlin,
381 F.3d at 1259;
see also Harriscom Svenska AB v. Harris Corp.,
947 F.2d 627, 631 (2d Cir. 1991) (“Where, as here, the controlling issues are questions of fact, or, more precisely, questions as to whether genuine issues of material fact remain to be tried, the federal scheme does not provide for an immediate appeal ....”). Significantly, there is “a distinction between a question of law, which will satisfy § 1292(b), and a question of fact or matter for the discretion of the trial court.”
See McFarlin,
381 F.3d at 1258 (internal quotation marks omitted).
In its statistical sampling ruling, the district court determined that the use of statistical sampling evidence can sometimes be permissible, but is not appropriate here based on the particular facts and evidence in this case. Moreover, in their opening appellate brief, the relators clarify that “[t]he true question for the District Court is
not
whether statistical sampling and extrapolation, in and of itself, is appropriate.”
See
Br. of Relators 11 (emphasis added). Rather, the relators insist that the issue is whether their proposed “statistical sampling is conducted in a scientifically proven and accepted manner pursuant to the Supreme Court’s ruling in
[Daubert v. Merrell Dow Pharmaceuticals, Inc.,
509 U.S. 579, 113 S.Ct. 2786, 125 L.Ed.2d 469 (1993) ].”
Id.
Thus, the relators’ appeal raises the question of whether the district court may, in its discretion, allow the rela-tors to use statistical sampling to prove their case.
See Bryte v. Am. Household, Inc.,
429 F.3d 469, 475 (4th Cir. 2005) (“[T]he district court has broad latitude in ruling on the admissibility of evidence, including expert opinion, and we will not overturn
Daubert
evidentiary rulings with respect to relevance and reliability absent an abuse of discretion.”).
In these circumstances, we are satisfied that, as to the statistical sampling ruling, the relators’ appeal does not present a pure question of law that is subject to our interlocutory review under § 1292(b). Accordingly, although we understand and appreciate the district court’s desire to obtain review of its statistical sampling ruling prior to undertaking complex trial proceedings, we are constrained to dismiss that aspect of the relators’ appeal as improvidently granted.
III.
Pursuant to the foregoing, we affirm the district court’s unreviewable veto ruling and dismiss as improvidently granted the relators’ appeal as to the court’s statistical sampling ruling.
AFFIRMED IN PART AND DISMISSED IN PART