PLANNED PARENTHOOD OF ARKANSAS & EASTERN OKLAHOMA, on behalf of itself and its patients, doing business as Planned Parenthood Great Plains; Stephanie Ho, MD, on behalf of herself and her patients, Plaintiffs-Appellees v. Larry JEGLEY, Prosecuting Attorney for Pulaski County, in his official capacity, his agents and successors; Matt Durrett, Prosecuting Attorney for Washington County, in his official capacity, his agents and successors, Defendants-Appellants. American Public Health Association; American College of Obstetricians and Gynecologists, Amici on Behalf of Appellee(s), 864 F.3d 953

GRUENDER, Circuit Judge. Prosecuting Attorneys for Pulaski County and Washington County, Arkansas (“the State”) appeal the district court’s grant of a preliminary injunction preventing the enforcement of an Arkansas statute requiring medication-abortion providers to contract with a physician who has hospital admitting privileges. Because the district court failed to make factual findings estimating the number of women burdened by the statute, we vacate the preliminary injunction and remand for further proceedings. I. BACKGROUND In 2015, Arkansas enacted the Abortion-Inducing Drugs Safety Act (“the Act”). Ark.

Code Ann. §§ 20-16-1501-1510. The Arkansas Legislature made findings that abortion-inducing drugs present significant medical risks, including “abdominal pain, cramping, vomiting, headache, fatigue, uterine hemorrhage, viral infections, and pelvic inflammatory disease.” Id. § 1502(14).

It further determined that medication abortions are “associated with an increased risk of complications relative to surgical abortion[s]” and found that, based on a 2011 United States Food and Drug Administration report, complications included eight deaths attributed to severe bacterial infection, 612 hospitalizations, 339 blood transfusions, and 256 infections. Id. §§ 1502(15)-(17). To address these health concerns, the Act created new requirements for physicians providing medication abortions.

Section 1504(

d) sets forth the “contract-physician requirement,” which is the subject of the current appeal. The provision requires that:

(1) The physician-who gives, sells, dispenses, administers, or otherwise provides or prescribes the abortion-inducing drug shall have a signed contract with a physician who agrees to handle complications and be able to produce that signed contract on demand by the patient or by the Department of Health.

(2) The physician who contracts to handle emergencies shall have active admitting - privileges and gynecological/surgical privileges at a hospital designated to handle any emergencies associated with the use or ingestion of the abortion-inducing drug.

(3) Every pregnant woman to whom a physician gives, sells, dispenses, administers, or otherwise provides or prescribes any abortion-inducing drug shall receive the name and phone number of the contracted physician and the hospital at which that physician maintains admitting privileges and which can handle any emergencies. Id. § 1504(d). The Act imposes civil and criminal penalties for violations, of the contract-physician requirement. See id. §§ 1506-Í507.

Planned Parenthood of Arkansas & Eastern Oklahoma (“PPAEO”) provides medication abortions in Arkansas at its two facilities, one in Fayetteville and the other in Little Rock.

The only other Arkansas abortion provider, Little Rock Family Planning Services (“LRFP”), administers both medication and surgical abortions at its Little Rock facility, PPAEO and one of its physicians, Stephanie Ho, M.D., (collectively “Planned Parenthood”) filed suit seeking to enjoin enforcement of the Act days before it was set to take effect, claiming that the contract-physician requirement unduly burdens their patients’ right to an abortion.

Both parties submitted affidavits concerning the medical benefits of the contract-physician requirement and the burdens on abortion access purportedly caused by the requirement. The district court found that Planned Parenthood’s protocols provided continuity of care because patients with concerns could call Planned Parenthood’s twenty-four-hour hotline to speak with nurses, Planned Parenthood referred patients experiencing complications to clinics or health centers for surgical completion, and Planned Parenthood physicians could consult with emergency-room physicians in the case of serious complications.

The district court thus concluded that the contract-physician requirement provided few, if any, tangible medical benefits over Planned Parenthood’s continuity-of-care protocols such that “the [Sjtate’s overall interest in the regulation of medication abortions through the [contract-physician] requirement is low and not compelling.” Planned Parenthood Ark. & E. Okla. v. Jegley, No. 4:15-cv-00784-KGB, 2016 WL 6211310, at *20 (E.D. Ark. Mar. 14, 2016). The district court then turned to the requirement’s alleged burdens on abortion access.

The court first concluded that Planned Parenthood could not find a physician to contract with and that, as a result, the Planned Parenthood facilities in Little Rock and Fayetteville would stop offering abortion services. It also found that medication abortion would no longer exist in Arkansas and that LRFP would be the sole abortion provider in Arkansas and would only administer surgical abortions.

The district court and the parties generally treated LRFP’s surgical-abortion services as a viable alternative to medication abortions, and as a result, the court determined the contract-physician requirement would not burden most Arkansas women seeking medication abortions because they already would have traveled to Little Rock prior to the enactment of the contract-physician requirement. The district court, however, found that the closure of PPAEO’s Fayetteville facility would force “women in the Fayetteville area” to make two, 380-mile round trips to obtain an abortion at LRFP. Id. at *4.

As a result of the increased travel distances, the district court determined that “some women” in the Fayetteville area would postpone the procedures, leading to an increased risk of complications, while others would forgo abortions entirely. Id. at *8.

The court further noted that the record did not allow a finding as to whether LRFP would be able to “absorb such an increase in the number of procedures or whether [LRFP] will be able to cover fully the needs of women who might have sought care at [Planned Parenthood].” Id. at *30. • Balancing the benefits of the contract-physician requirement against its burdens, the district court concluded that the requirement was a “solution in search of a problem.” Id. at *18.

It thus held that Planned Parenthood was likely to succeed on the merits, that it and its patients faced irreparable harm, that the equities weighed in its favor, and that the public interest weighed in its favor. As a result, the district court granted Planned Parenthood '.a preliminary injunction, preventing Arkansas from enforcing the contract-physician requirement. The State timely appealed. II. DISCUSSION This court has jurisdiction under 28 U.S.C. § 1292(a)(1) to review an interlocutory order granting a preliminary injunction. We review such an order for an abuse of discretion.

Planned Parenthood Minn., N.D., S.D. v. Rounds, 530 F.3d 724, 733 (8th Cir. 2008) (en banc). A district court abuses its discretion when it fails to consider a relevant factor that should have been given significant weight, when it considers and gives significant weight to an irrelevant or improper .factor, or when it considers only proper, factors—and no improper ones—but in weighing those factors commits a clear error of judgment. Novus Franchising, Inc. v. Dawson, 725 F.3d 885, 893 (8th Cir. 2013).

Generally, in is'suing a preliminary injunction, the district court considers: (1) the threat of irreparable harm to the moving party, (2) the balance between this harm and the injury that granting the injunction will inflict on the non-moving party, (3) the probability that the moving party will succeed on the merits, and (4) the public interest. See Dataphase Sys., Inc. v. C L Sys., Inc., 640 F.2d 109, 114 (8th Cir. 1981) (en banc).

Where a preliminary injunction is sought to enjoin the implementation of a duly enacted state statute, however, the moving party must make a more rigorous showing that it is “likely to prevail on the merits.” Rounds, 530 F.3d at 732-33. This is necessary “to ensure that preliminary injunctions that thwart a state’s presumptively reasonable democratic processes are pronounced only after an appropriately deferential analysis.” Id. at 733. Thus, we must analyze whether Planned Parenthood demonstrated that it is likely to prevail on the merits of its undue burden claim.

See id. at 732. “A finding of an undue burden is a shorthand for the conclusion that a state regulation has the purpose or effect of placing a substantial obstacle in the path of a woman seeking an abortion of a nonviable fetus.” Planned Parenthood of Se. Pa. v. Casey, 505 U.S. 833, 877, 112 S.Ct. 2791, 120 L.Ed.2d 674 (1992). In Whole Woman’s Health v. Hellerstedt, the Supreme Court clarified that this undue burden analysis “requires that courts consider the burdens a law imposes on abortion access together with the benefits those laws confer.” — U.S. —, 136 S.Ct. 2292, 2309, 195 L.Ed.2d 665 (2016).

The Court explained that after the passage of Texas House Bill 2 (“H.B. 2”), the abortion regulation at issue, the number of Texas facilities providing abortions decreased from approximately forty to about seven or eight. Id. at 2312, 2316. These closures led to increased driving distances, though the additional driving distances alone were not dispositive. Id. at 2313 (“We recognize that increased driving distances do not always constitute an ‘undue burden.’ ” (citing Casey, 505 U.S. at 885-87, 112 S.Ct. 2791)).

Instead, the closures burdened abortion access because women seeking abortions also faced “fewer doctors, longer waiting times, and increased crowding.” Id. Furthermore, patients would be “less likely to get the kind of individualized attention, serious conversation, and emotional support” at the abortion facilities. Id. at 2318. As a result, the Supreme Court struck down H.B. 2 because its numerous burdens substantially outweighed its benefits. See id. at 2313, 2318. At the same time, because Hellerstedt expressly relied on Gonzales v.

Carhart, see id. at 2310, the Court preserved its command that “state and federal legislatures [have] wide discretion to pass legislation in areas where there is medical and scientific uncertainty,” 550 U.S. 124, 163, 127 S.Ct. 1610, 167 L.Ed.2d 480 (2007). In the present case, the district court abused its discretion because it failed to consider whether Planned Parenthood satisfied the requirements necessary to sustain a facial challenge to an abortion regulation. “Facial challenges are disfavored,” Wash. State Grange v. Wash.

State Republican Party, 552 U.S. 442, 449, 128 S.Ct. 1184, 170 L.Ed.2d 151 (2008), and generally, they can only succeed if the proponent establishes that “no set of circumstances exists under which the [statute] would be valid,” United States v. Salerno, 481 U.S. 739, 745, 107 S.Ct. 2095, 95 L.Ed.2d 697 (1987).

For challenges to abortion regulations, however, the Supreme Court has fashioned a different standard under which the plaintiff can prevail by demonstrating that “in a large fraction of the cases in which [the law] is relevant, it will operate as a substantial obstacle to a woman’s choice to undergo an abortion.” Casey, 505 U.S. at 895, 112 S.Ct. 2791.

The Supreme Court has clarified that “cases in which the provision at issue is relevant” is a narrower category than “all women,” “pregnant women,” or even “women seeking abortions identified by the State.” Hellerstedt, 136 S.Ct. at 2320 (quoting Casey, 505 U.S. at 894-95, 112 S.Ct. 2791). Thus, because the contract-physician requirement only applies to medication-abortion providers, the “relevant denominator” here is women seeking medication abortions in Arkansas.

See id. (finding that the “relevant denominator” must be “those women for whom the provision is an actual rather than an irrelevant restriction” (internal alterations omitted)). Accordingly, in order to sustain a facial challenge and grant a preliminary injunction, the district court was required to make a finding that the Act’s contract-physician requirement is an. undue burden for a large fraction of women seeking medication abortions in Arkansas.

The district court did not make this finding., The court correctly held that individuals for whom the contract-physician requirement was an actual, rather than an irrelevant, restriction were women seeking medication abortions in Arkansas. Nonetheless, it did not define or estimate the number of women who would be unduly burdened by the contract-physician requirement. Instead, it focused on amorphous groups of women to reach its conclusion that the Act was facially unconstitutional. First, the district court did not determine how many 'women would face increased travel distances.

The court noted that most women residing in Arkansas and seeking medication abortions would be unaffected by the contract-physician requirement, as they could travel to LRFP for an abortion. However, it found that “women in the Fayetteville area” would have to make two, 380-mile round trips to obtain an abortion from LRFP in Little Rock. Jegley, 2016 WL 6211310, at *4. Nonetheless, it is unclear how many women would have to travel these additional distances.

For example, the district court did not explain if “women in -the Fayette-ville area” referred to women residing only in the city of Fayetteville, women residing in Washington County (where Fayettéville is located), or women residing in surrounding counties as well. Additionally, as the Supreme Court acknowledged in Heller-stedt, increased travel distances are relevant but may not independently constitute an undue burden. 136 S.Ct. at 2313 (citing Casey, 505 U.S. at 885-87, 112 S.Ct. 2791).

The Supreme Court found an undue burden in Hellerstedt because women seeking abortions faced “fewer doctors, longer waiting times, and increased crowding.” 136 S.Ct. at 2313. Here, it is not clear that “women in the Fayetteville area” traveling to LRFP would face “fewer doctors, longer waiting times, and increased crowding.” See id.

As the district court recognized, the record did not demonstrate whether LRFP would be able to “absorb such an increase in the number of procedures or whether [LRFP] [would] be able to cover fully the needs of women who might have sought care at [Planned Parenthood].” Jegley, 2016 WL 6211310, at *30. Next, the district court failed to estimate the number of women who would forgo abortions. The court cited an affidavit from Dr. Stanley K.

Henshaw, Ph.D., who opined that an increased travel distance of 100 miles would cause 20 to 25 percent of women who would have otherwise obtained abortions to forgo them and that “[g]reater distances will be a barrier to an even higher percentage of women.” The record is unclear as to whether the 100 miles of increased travel distance refers to round-trip or one-way distances—or whether it concerns single or multiple trips.

More fundamentally, however, the district court did not apply this conclusion to estimate the number of women in the Fayetteville area seeking medication abortions who would actually forgo abortions. Finally, the court did not estimate the number of women who would postpone their abortions, The district court maintained that increased travel distances would cause “some women” in the Fayette-ville area to postpone their abortions' and thereby face an increased risk of complications. Id. at *8.

The district court again, however, did not explain or estimate how many women constituted “some women.” While the record does indicate that delaying abortions can increase the risk of complications, the court failed to estimate the number of women who would face ah increased risk of complications, As a result, we are left with no concrete district court findings estimating the, number of women who would be unduly burdened by the contract-physician requirement—either because they would forgo the procedure or postpone it—and whether they constitute a “large fraction” of women seeking medication abortions in Arkansas such that Planned Parenthood could prevail in its facial challenge to the contract-physician requirement.

In situations'like this, where the district court did not make the necessary factual findings, “[w]e conclude that the better course is to afford the district court an opportunity to make appropriate findings of fact and con-elusions of law.” See Phelps-Roper v. Troutman, 712 F.3d 412, 417 (8th Cir. 2013) (per curiam); see also Mo. Pac. Joint Protective Bd., Bhd. Ry. Carmen v. Mo. Pac. R.R. Co., 730 F.2d 533, 537 (8th Cir. 1984) (“[W]e believe the findings and conclusions should, in the first instance, be made by the district court.”).

On remand; we do not require the district court to calculate the exact number of women unduly burdened by the contract-physician requirement. We acknowledge that-the “large fraction” standard is in some ways “more conceptual than mathematical.” Cincinnati Women’s Servs., Inc. v. Taft, 468 F.3d 361, 374 (6th Cir. 2006). Nonetheless, like the Sixth Circuit; we find that this standard is not entirely freewheeling and that we can and should define its outer boundaries.

See id. (“[T]he term ‘large fraction,’ which, in a way, is more conceptual than mathematical, envisions, something more than the 12 out of 100 women identified here.”). Thus, on remand, the district court should conduct fact finding concerning the number of women unduly burdened by the contract-physician requirement and determine whether that number constitutes a “large fraction.” Accordingly, we vacate the district court’s grant of a preliminary injunction and remand for further proceedings consistent with this opinion. .

The Act also requires physicians administering medication abortions to follow an FDA-approved regimen, which differed from the one Planned Parenthood used. The district court enjoined this portion of the. Act along with the contract-physician requirement. Following the issuance of the preliminary injunction,, the FDA updated its regimen to one that reflected Planned Parenthood's regimen. As a result, Planned Parenthood withdrew its challenge to this provision, and, thus, the requirement that physicians follow FDA regulations is not before us. .

Planned Parenthood’s efforts to recruit a contract physician did not include any offer of financial compensation. It is unclear whether the district court considered this fact in its assessment. . The court noted that medication'abortions could be medically indicated for women with specific health conditions. However, it also acknowledged that the record was "unclear" as to "what percentage of the patient population that may be.” Jegley, 2016 WL 621131Ó, at *30. . 'A separate Arkansas statute requires women to receive state-mandated information forty-eight hours before their abortion procedure. See■ Ark.

Code Ann. § 20-16-1703(b)(l), (2), . This information must be given "orally and in person,” thereby possibly necessitating another trip. Id. . Indeed, in 2014, medication abortions accounted for only 14.3 percent of all abortions in Arkansas. . Although the record does contain evidence that, in 2014, 145 women residing in Washington County had medication abortions, applying the 20 to '25 percent figure would mean that about 29 to 37 women would forgo their abortions—approximately 4.8 to 6.0 percent of all medication abortions provided in Arkansas in 2014.

We are skeptical that 4.8 to 6.0 percent is sufficient to qualify as a "large fraction” of women seeking medication abortions in Arkansas. See Cincinnati Women's Servs., Inc. v. Taft, 468 F.3d 361, 374 (6th Cir. 2006) (holding that 12 percent does not constitute a "large fraction”), , We find it unnecessary to reach the issue of the contract-physician requirement’s benefits, though the district court's method gives us some pause.

In determining that the contract-physician requirement's benefits would be "low and not compelling,” the district court concluded that Planned Parenthood’s current continuity-of-care protocols were adequate. Hellerstedt, however, compared H.B. 2 to Texas's pre-existing law, not Texas abortion providers’ current protocols, See 136 S.Ct, at 2311, ("We have found nothing in Texas' record evidence that shows that, .compared to prior law (which required a 'working arrangement’ with, a doctor with admitting privileges), the new law advanced Texas’ legitimate interest in protecting women's health.” (emphasis added)).

Moreover, Planned Parenthood could unilaterally decide to discontinue its twenty-four-hour nurse-staffed phone line, end patient referrals to surgical providers, o? stop, consultations with emergency-room physicians in the case of serious complications, While we elect not to quantify it at this time, we certainly see some benefit for patients where the State mandates continuity-of-care standards—especially in the face of known complications and where there previously had been no state requirements.

For instance, had the State merely, mandated Planned Parenthood's existing continuity-of-care- protocols, Planned Parenthood likely would not argue that these would be of no significant, benefit to'its patients. At' the very least, codifying Planned Parenthood’s continuity-of-care protocols would constitute a benefit because it would set a legal floor to prevent retrenchment in the standard of care. The question here, however, is whether the contract-physician requirement’s benefits are substantially outweighed by the burdens it imposes on a large fraction of women seeking medication abortion in Arkansas.