PER CURIAM:

In this expedited appeal, Assistant Secretary of Labor Thorne G. Auchter challenges a January 5, 1983, district court order, 554 F.Supp. 242, order directing the Occupational Safety and Health Administration (OSHA) to issue by January 25, 1983, an emergency temporary standard (ETS) regulating workplace exposure to ethylene oxide (EtO), a synthetic organic chemical. Used, inter alia, as a sterilizing agent, fumigant, pesticide, and industrial chemical additive, EtO is critically employed as a sterilant in the health care and medical products industries. The exposure of workers who work near sterilizing equipment in the health care industry is the principal focus of this case.

The current OSHA standard for EtO has been in effect for well over a decade; it allows a permissible exposure limit in the workplace of 50 parts per million (ppm) averaged over an eight hour workday ("time-weighted average" or TWA). 29 C.F.R. § 1910.1000. That level, information now available reveals, poses a serious risk of causing chromosomal abnormalities and cancer in exposed animals and humans.

In August 1981, Public Citizen Health Research Group ("Public Citizen") petitioned OSHA to issue an emergency temporary standard reducing the permissible exposure limit of EtO to a TWA of 1 ppm. Public Citizen invoked a provision of the Occupational Safety and Health Act, 29 U.S.C. §§ 651-78 (1976 and Supp. IV 1980) (OSH Act or Act), that authorizes OSHA to adopt an immediately effective temporary standard upon determining that failure to take emergency action would subject employees to "grave danger" from exposure to a toxic substance. 29 U.S.C. § 655(c). In September 1981, OSHA denied the petition on the ground that "on the available evidence . . . current conditions do not constitute an emergency situation." App. 5a. The evidence did persuade the Assistant Secretary, however, that "the current OSHA standard of 50 ppm may not be sufficiently protective"; he therefore stated he would proceed with normal non-emergency rulemaking to revise the standard. App. 5b; Brief for the Appellants at 4, 25 (hereafter referred to as OSHA Brief).

Four months later, in January 1982, OSHA published an "Advance Notice of Proposed Rulemaking." 47 Fed.Reg. 3566 (Jan. 26, 1982); App. 170-75. The Assistant Secretary projects issuance of an actual Notice of Proposed Rulemaking in June 1983, and a final regulation sometime in the fall of 1984.

Shortly before OSHA denied its petition, Public Citizen commenced this civil action to obtain an order commanding the issuance of an emergency standard. On cross-motions for summary judgment, the district court concluded that "[t]he record before the agency presented a solid and certain foundation showing that workers are subjected to grave health dangers from exposure to ethylene oxide within the currently permissible range." App. 141. OSHA had abused its discretion, the district court held, by denying the Public Citizen petition and, instead, proceeding on a course "which insured the continuing existence of the challenged standard." Id. Accordingly, the district court ordered OSHA to "promulgate within 20 days . . . an appropriate emergency temporary standard addressing worker exposure to ethylene oxide." App. 146.

This difficult case, which we must decide under pressing circumstances, has two novel aspects. First, OSHA has in the past, on its own initiative, issued emergency standards which were promptly brought to court for review by industry complainants. See Florida Peach Growers Association, Inc. v. United States Department of Labor, 489 F.2d 120 (5th Cir. 1974); Dry Color Manufacturers' Association, Inc. v. United States Department of Labor, 486 F.2d 98 (3d Cir. 1973) (both vacating emergency temporary standards). We are unaware, as was the district court, App. 134, of any prior case in which a court was invited to review OSHA's denial of a petition to issue an emergency standard. Second, in declining to replace the current standard on an emergency basis, and in justifying its position in court, OSHA did not rely on any government action. Rather, it relied heavily on voluntary efforts of employers, "alerted and responsive to the new health data concerning [EtO]," to "lower their in-house allowable exposure limits to a fraction of the OSHA standard." OSHA Brief at 22, 23.

The Assistant Secretary has emphasized the extraordinary authority and large measure of discretion 29 U.S.C. § 655(c) vests in him to determine whether an "emergency standard is necessary to protect employees from [grave] danger." While it is a close question, our review of the record indicates that, in ordering an emergency standard, the most drastic measure in the Agency's standard-setting arsenal, the district court impermissibly substituted its evaluation for that of OSHA. Nonetheless, we fully agree with the district court that "OSHA has embarked upon the least responsive course short of inaction." App. 135. Beyond question, despite the efforts of "[m]any companies," OSHA Brief at 23, the record shows a significant risk that some workers, and the children they will hereafter conceive, are subject to grave danger from the employees' exposure to EtO. We therefore hold that OSHA must expedite the rulemaking in which it is now engaged.

Congress has instructed OSHA, in determining the priority for establishing standards, to "give due regard to the urgency of the need for mandatory . . . health standards for particular . . . workplaces." 29 U.S.C. § 655(g). Further, the Administrative Procedure Act directs an agency "to conclude [within a reasonable time] a matter presented to it." 5 U.S.C. § 555(b). Complementing that provision, this court has authority, pursuant to 5 U.S.C. § 706(1), to compel agency action unreasonably delayed. There is an obvious need, apparent to OSHA, for an EtO standard that reflects, as the current standard does not, the mutagenic and carcinogenic potential of the chemical. Because it is also plain that industry efforts fall a considerable distance from taking every worker exposed to EtO out of the grave danger zone, a more than three-year span from Public Citizen's petition to projected final regulation is not tolerable. OSHA's failure to date to issue even a notice of proposed rulemaking — some eighteen months after announcing its intention to commence rulemaking — is in our judgment, and in light of the risk to current and future lives, agency action unreasonably delayed.

To assure that OSHA will give due regard to the need, urgent for some workers, for a new EtO standard, and to prevent undue protraction in OSHA's conclusion of this matter, we direct the Assistant Secretary to issue a notice of proposed rulemaking within thirty days of the date of this decision and to proceed expeditiously thereafter toward issuance of a permanent standard for EtO.

I. BACKGROUND

It is undisputed that evidence now available shows EtO to pose serious qualitative health risks unappreciated when OSHA adopted the current standard. The chemical is both mutagenic and carcinogenic in animals and humans. One uncontradicted study finds significant chromosomal aberrations in workers chemically exposed to 36 ppm, App. 996-1003; another study shows similar results for concentrations described, more generally, as within the current 50 ppm standard. App. 1488-99. In addition, dose responsive relationships between exposure to EtO and fertility in rats have been observed. 47 Fed.Reg. 3568 (Jan. 26, 1982); App. 172.

OSHA concedes that scientific reports show a "statistically significant" increase in cancer in animals exposed at 33 ppm and above. The National Institute of Occupational Safety and Health (NIOSH) determined from the prior reports that "exposures as low as 10 ppm increased the proportion of female rats with . . . leukemia." App. 954. OSHA has acknowledged this dose responsive relationship at 10 ppm. 47 Fed.Reg. 3568 (Jan. 26, 1982); App. 172.

The two principal human studies reported a "significant excess" of both leukemia and stomach cancer from exposure to EtO, OSHA Brief at 15; see also 47 Fed.Reg. 3568; App. 172; the same cancers appeared in animals exposed to EtO. The human studies indicated a 15-fold increase in leukemia at 20 ± 10 ppm and a statistically significant increase in total cancer mortality at exposures averaging well within the current 50 ppm standard. 47 Fed.Reg. 3568; App. 172, 1041-42, 1053-57.

The evidence is more problematic on the quantitative risk posed by EtO, i.e., the extent to which workers are actually exposed to the serious qualitative risks of the chemical. NIOSH estimates that 75,000 health care workers are regularly exposed to EtO and another 25,000 "casually" or intermittently exposed. App. 898. But uncontradicted evidence in the record indicates that many hospitals have voluntarily improved internal procedures or modified their sterilizers to reduce exposure levels to 10 ppm and below. See, e.g., App. 839-44. Other evidence presented to the district court, however, suggests that a not insignificant number of workers are still regularly exposed to levels much higher than 10 ppm, but within the current 50 ppm standard. A 1977 study by the American Hospital Association (AHA) reported that although 24 of the 27 hospitals studied had chronic daily exposure levels of 10 ppm and below, the remaining three had chronic exposures of 50 ppm and above. App. 821-84. A 1981 MITRE Corporation study, incorporating the AHA 1977 study results, found of 65 hospitals examined, 24 did not even monitor EtO levels. App. 523.

Evaluating this evidence, the Assistant Secretary acknowledged that "there is important new health data on the carcinogenic risks of EtO," but determined that at current levels of exposure EtO did not present a danger sufficiently grave to require issuance of an ETS. App. 5a. The Agency initially stated:

[t]he limited data currently available to OSHA . . . indicate that most hospitals have reduced exposures below 10 ppm TWA for sterilizing operations. The data also suggest that the number of employees at any given hospital who may be involved in using EtO as a sterilant is very small, and that their exposure is intermittent.

App. 5b. During litigation, OSHA slightly qualified or clarified this assessment and estimated the "average" hospital exposure level at (neither above nor below) 10 ppm.

Public Citizen disputes OSHA's findings as unsupported by the record. Further, Public Citizen asserts that even if the "average" exposure level is indeed 10 ppm and even if the risk of harm at that level is not "grave," the record still shows many workers exposed to levels over 10 ppm, yet below the current standard of 50 ppm, levels at which the risk of harm is grave.

II. DECISION

A. Standard of Review

The legislative history of 29 U.S.C. § 655(c) makes clear that emergency standards are to be used only in "limited situations." Correspondingly, courts have observed that authority to adopt such standards entails "[e]xtraordinary power," Florida Peach Growers Association, Inc. v. United States Department of Labor, 489 F.2d 120, 129 (5th Cir. 1974), that should be "delicately exercised," id., and used only as "an unusual response to exceptional circumstances." Dry Color Manufacturers' Association, Inc. v. United States Department of Labor, 486 F.2d 98, 104 n. 9a (3d Cir. 1973); see also Taylor Diving Salvage v. United States Department of Labor, 537 F.2d 819, 820-21 (5th Cir. 1976).

In deciding whether to issue an ETS, OSHA must make both factual and policy judgments on the basis of information that may be incomplete. The Supreme Court, in the context of permanent regulation, has noted that threshold determinations by OSHA of whether a sufficient risk of harm exists to justify regulation "must necessarily be based on considerations of policy as well as empirically verifiable facts." Industrial Union Department v. American Petroleum Institute, 448 U.S. 607, 655 n. 62, 100 S.Ct. 2844, 2871 n. 62, 65 L.Ed.2d 1010 (1980) ( Benzene). While OSHA determinations "which are essentially legislative and rooted in inferences from complex scientific and factual data," United Steelworkers of America v. Marshall, 647 F.2d 1189, 1206 (D.C. Cir. 1980), cert. denied sub nom. Lead Industries Association, Inc. v. Donovan, 453 U.S. 913, 101 S.Ct. 3148, 69 L.Ed.2d 997 (1981), are entitled to great deference, our review of the Assistant Secretary's refusal to issue an ETS for EtO exposure must take into account the mandatory language of 29 U.S.C. § 655(c) and the fact that "the interests at stake are not merely economic interests in a license or a rate structure, but personal interests in life and health." Wellford v. Ruckelshaus, 439 F.2d 598, 601 (D.C. Cir. 1971); see Nader v. Nuclear Regulatory Commission, 513 F.2d 1045, 1047 (D.C. Cir. 1975).

B. Merits 1. OSHA's finding of no "grave danger" is rational.

The district court, after reviewing the available evidence, concluded that "the agency's decision resulted from a clear error of judgment." App. 135. Responding to OSHA's position that there was no emergency because private institutions had voluntarily reduced EtO exposure to levels OSHA considered "safe," the district court declared "[s]uch actions do nothing to protect those workers who are still exposed to EtO at hazardous level or those employed at institutions which may at any time elect to withdraw the voluntary restrictions." App. 137 (emphasis added). The district court thus cited actual exposure risks and not, as OSHA would restate the court's position, simply a future prospect should employers relax their vigilance.

In light of the mixed fact/policy judgment Congress empowered OSHA to make on uncertain evidence, we cannot say, as the district court did, that the decision not to issue an ETS lacked support in the record. The Agency's estimates of a 10 ppm average exposure level and a statistically lower risk of harm at that level may rest on impressionistic information. However, nothing offered to the district court contradicts those estimates. Given the substantial evidence of voluntary reduction in exposure levels and the predominantly lower levels reported in the available studies, a 10 ppm "average" appears to us a reasonable assumption. Similarly, the absence of clear scientific evidence of the degree of harm from exposure at levels 10 ppm and below counsels deference to the Assistant Secretary's assessment.

The difficult question in this case is whether the Assistant Secretary's reliance on an average exposure level in finding no grave danger was irrational or an abuse of discretion. We are not positioned to say that the expert Agency acted impermissibly in this regard. From the statistically small sampling and the paucity of current data concerning actual EtO exposure levels, we are unable to venture even a guess as to existing exposure patterns over the average. Under the circumstances, we must defer to the Assistant Secretary's determination that the available evidence is inadequate to show the existence of "grave danger" necessitating an emergency standard.

All we say today is that in the absence of a more complete record as to actual exposure levels, we are hesitant to compel the Assistant Secretary to grant extraordinary relief. We express no opinion as to whether the same record would support voluntary issuance by OSHA of an emergency standard.

2. The significant risk of grave danger necessitates expedited rulemaking.

Though we disagree with the district court that a finding of grave danger is compelled on this record such that the Assistant Secretary should be required to exercise his "extraordinary power" to direct immediate relief in the form of an ETS, our review would be less than "thorough" and "probing" if we simply overturned the judgment below. Ample evidence in the record indicates a significant risk that some workers, who are actually being exposed to levels of EtO greater than the 10 ppm "average" (yet within the 50 ppm standard), currently encounter a potentially grave danger to both their health and the health of their progeny. In face of this evidence, the Assistant Secretary's unaccounted-for delay in issuing a Notice of Proposed Rulemaking and his refusal to assign to the EtO rulemaking any priority status constitute agency action "unreasonably delayed." 5 U.S.C. § 706(1).

In its September 1981 denial of Public Citizen's petition for an ETS, OSHA conceded that its current 50 ppm standard "may not be sufficiently protective," and stated its "intent to proceed with rulemaking under § 6(b) of the OSH Act [ 29 U.S.C. § 655(b)] to lower the permissible exposure limit and to incorporate appropriate provisions for monitoring, medical surveillance and the like." App. 5b. The Assistant Secretary encouraged Public Citizen "to provide information in response to OSHA's advance notice of proposed rulemaking, and to participate in the full rulemaking proceedings on EtO." Id. Despite this announced intent in September 1981 to proceed with a rulemaking, no notice of proposed rulemaking has yet been issued.

Instead, the Assistant Secretary issued in January 1982, four months after committing OSHA to a rulemaking, merely an "Advance Notice of Proposed Rulemaking". See 47 Fed.Reg. 3566 (Jan. 26, 1982); App. 170-75. The comment period on this "advance notice" ended March 31, 1982. Almost a year has passed without commencement of the rulemaking. The Assistant Secretary now estimates that a notice of proposed rulemaking will issue in June 1983, with promulgation of a final rule sometime in the fall of 1984.

Three years from announced intent to regulate to final rule is simply too long given the significant risk of grave danger EtO poses to the lives of current workers and the lives and well-being of their offspring. Delays that might be altogether reasonable in the sphere of economic regulation are less tolerable when human lives are at stake. See, e.g., Blankenship v. Secretary of Health, Education, and Welfare, 587 F.2d 329, 334 (6th Cir. 1978); Environmental Defense Fund, Inc. v. Hardin, 428 F.2d 1093, 1099 (D.C. Cir. 1970). This is particularly true when the very purpose of the governing Act is to protect those lives. 29 U.S.C. § 651(b).

We would hesitate to require the Assistant Secretary to expedite the EtO rulemaking if such a command would seriously disrupt other rulemakings of higher or competing priority. But we do not confront such a case. Prior to oral argument, we asked the parties to address the status of other ongoing OSHA rulemakings and the feasibility of expediting the EtO proceedings in the absence of an ETS. OSHA informed us that the Agency is currently engaged in three proceedings that it believes would be disturbed by speedier EtO rulemaking: (1) rulemaking requiring chemical manufacturers to assess and communicate to employees the hazards posed by chemicals they produce; (2) a reassessment of the current regulations governing worker exposure to asbestos; and (3) rulemaking governing exposure to ethylene dibromide.

The hazard communication rulemaking has been on the Agency's docket since 1977, see 42 Fed.Reg. 5372 (Jan. 28, 1977); by OSHA's own admission, development of the standard is nearly complete with a final rule expected by July 1983. 47 Fed.Reg. 48549 (Oct. 28, 1982). In addition to the fact that OSHA's work on the rule is substantially done, we note that this standard relates solely to information employers must give to workers concerning workplace hazards; it does not substantively regulate the workplace environment. The asbestos docket has been open since 1975, and OSHA estimates a Notice of Proposed Rulemaking may be issued in March 1984. Ethylene dibromide, a chemical with both carcinogenic and mutagenic risks, is appropriately on a somewhat faster track. An Advance Notice of Proposed Rulemaking issued in December 1981, and the Notice of Proposed Rulemaking is scheduled for May 1983. OSHA concedes, however, that virtually all exposure to ethylene dibromide takes place in California, which has recently issued an emergency standard setting the permissible exposure limit at 130 ppm. See 46 Fed.Reg. 61672 (Dec. 18, 1981); OSHA Brief at 57 n. 60. None of these proceedings appears to approach in urgency the need for prompt issuance of a new EtO exposure standard, and OSHA has provided us with no reasoned explanation why it has protracted the EtO rulemaking despite the documented risks to workers' lives and the lives of children they may conceive.

The court is not disarmed in these circumstances. OSHA is obliged by its governing Act to "give due regard to the urgency of the need" for a new EtO standard in setting priorities, and the record before us strongly indicates that the Agency has not done so. OSHA, we are convinced, is not proceeding to conclude this matter "within a reasonable time." 5 U.S.C. § 555(b). We therefore act to "compel . . . action" OSHA has "unreasonably delayed." 5 U.S.C. § 706(1).

CONCLUSION

We cannot "compel solutions where none exist," American Broadcasting Co. v. FCC, 191 F.2d 492, 501 (D.C. Cir. 1951), but we "must act to make certain that what can be done is done." Id. To that end, we order the Assistant Secretary to issue a notice of proposed rulemaking within 30 days of the date of this decision. Although we dictate no fixed date for issuance of a final rule, we do direct OSHA to proceed on a priority, expedited basis and to issue a permanent standard as promptly as possible, well in advance of the current, latter part of 1984 estimate. Under the circumstances presented here, i.e., the significant risk of grave danger to human life, and the time OSHA has already devoted to EtO, we expect promulgation of a final rule within a year's time.

So Ordered.