James CAMPBELL, Petitioner-Respondent, v. NEW JERSEY RACING COMMISSION, Respondent-Appellant.
The opinion of the Court was delivered by
This appeal implicates the standards governing appellate review of administrative actions. Specifically, it involves the method by which the New Jersey Racing Commission (Commission) measures the level of total carbon dioxide (tCO2) in race horses. An abnormally high level of tCO2 is an indication that a horse has ingested or been injected with sodium bicarbonate or other substances to enhance the horse's performance during a race. That process of ingestion or injection is known in the racing industry as “milkshaking.”
In this case, the Commission found that a horse contained impermissible levels of tCO2 in its blood following a race at Freehold Raceway. Accordingly, the Commission assessed penalties against the horse's trainer, James Campbell (respondent), consistent with the agency's regulations. The Appellate Division set aside the Commission's actions. We granted the Commission's petition for certification, 167 N.J. 90, 769 A.2d 1052 (2001), and now reverse.
In accordance with its statutory authority, see N.J.S.A. 5:5-30, the Commission promulgated certain regulations setting forth the standards and procedures by which it would test for impermissible levels of tCO2 and assess penalties against any trainer of a horse found in that condition. Specifically, N.J.A.C. 13:71-23.1(b) states in relevant part:
On the day of the race, irrespective of the date, time and method of administration, no horse entered to start in or participating in any race shall carry in its body any drug and/or substance foreign to the natural horse ․ [including] ․ [s]odium bicarbonate (baking soda) ․ [and] ․ [m]ixtures, compounds or solutions commonly referred to as “milkshakes” which contain any prohibited drug and/or substance.
Further, N.J.A.C. 13:71-23.3A(a) states in relevant part:
An excess level of total carbon dioxide [tCO2] in the race horse is deemed adverse to the best interests of harness racing, and adverse to the best interests of the horse in that such condition alters its normal physiological state. Accordingly, ․ on the date of the race and following a minimum one-hour standing at rest period for the horse subsequent to the conclusion of the race within which it competed, a State Veterinarian representing the Commission may obtain blood samples from the horse for the purpose of the testing of said samples by the Racing Commission laboratory for [tCO2] level on a Clinical Auto Analyzer that applies an ion selective electrode method (ISE) for measuring [tCO2] in blood. Where the [tCO2] level, based upon such testing equals or exceeds the following levels, the judges shall order the relief authorized pursuant to (b) ․:
1. Thirty-seven (37) or more millimoles per liter[.]
As a penalty for a first violation, subsection (b) provides that the trainer, “as the absolute insurer of the horse[,]” shall have his or her license suspended for a 75-day period, be ordered to pay a $1000 fine, “and be denied the privileges of all grounds subject to the jurisdiction of the Commission during the suspension period.” N.J.A.C. 13:71-23.3A(b)1.
On October 25, 1997, Ramses Two (Ramses), a standard-bred horse trained by respondent, won a harness race at Freehold Raceway. After the race, and in accordance with the procedures set forth in the Commission's regulations, a track veterinarian extracted a blood sample from the horse to test for impermissible levels of tCO2. The test indicated a level of tCO2 in Ramses that exceeded the regulatory limit of 37 millimoles per liter (mmol/l). (A “mole” is a measure of molecular weight. A “millimole” is one one-thousandth of a mole.) The initial testing consisted of a screening test and a second confirmatory test, both of which detected excess levels of tCO2 in Ramses' blood.
Based on that initial testing, the Commission's board of judges charged respondent with a regulatory violation. Respondent disputed the charge. He invoked a provision of the Commission's procedures that requires a horse whose level of tCO2 has exceeded the permissible limit to be placed under guarded quarantine and re-tested to enable the animal's owner or trainer to demonstrate that such level is “physiologically normal for the particular horse[.]” N.J.A.C. 13:71-23.3A(c). The results of the re-test administered during the quarantine period confirmed that Ramses' previously-tested level of tCO2 was abnormal.
The Commission sanctioned respondent as a first-time violator. Respondent then requested a hearing before an administrative law judge (ALJ), and the matter proceeded as a contested case in accordance with statutory procedures. See N.J.S.A. 52:14B-1 to -15 and N.J.S.A. 52:14F-1 to -13.
The ALJ conducted a seven-day hearing during which the parties presented sharply divided expert testimony regarding the method by which the Commission tests equine plasma for impermissible levels of tCO2. That method involves a machine, the Beckman EL-ISE instrument, used by the Commission as the Clinical Auto Analyzer noted in N.J.A.C. 13:71-23.3A. Respondent does not challenge the use of the machine but rather the method by which the Commission calibrates the instrument before testing to verify accurate measurements.
The Commission's expert, Philip Lorimer, the Supervising Forensic Scientist for the New Jersey State Police, presented to the ALJ an overview of the milkshaking problem in North America. Lorimer explained that the introduction of bicarbonated substances into the bloodstream of a race horse theoretically improves its performance by increasing its blood's hydrogen ion concentration. He also stated that the Commission had arrived at the level of 37.0 mmol/l as its regulatory limit by testing the tCO2 levels in 231 randomly selected race horses in New Jersey. The mean tCO2 level for those horses was 30 mmol/l, and only one horse in the experiment had a reading higher than 34.0 mmol/l. Lorimer testified that the Commission had arrived at 37.0 mmol/l as the regulatory limit because it was more than three standard deviations above the mean that the Commission had discovered. Although some horses receive residual bicarbonates in their feed, Lorimer stated that the threshold value of 37.0 mmol/l is high enough to ensure that a “non-milkshaked” horse would not exceed it.
The Commission's expert also explained how the agency used the Beckman instrument. Lorimer stated that before testing a blood sample the machine is calibrated using solutions prepared by Beckman, the instrument's manufacturer. The calibration solutions contain tCO2 levels of 0, 10, and 30 mmol/l. The machine analyzes the content of the calibrating solutions to verify that its readings are proper. The manufacturer claims that the machine's readings are accurate up to 40 mmol/l if its calibration is checked properly. Lorimer further testified that to demonstrate quality control in the calibration, solutions from a different company (Casco solutions or standards) must be used to verify the calibration. Casco solutions of 10, 20, 30, and 40 mmol/l of tCO2 are run through the machine as a linearity check. Lorimer explained that the Casco 40 mmol/l solution was necessary to ensure the reliability of the blood sample measurements up to 40 mmol/l because the Beckman company did not make calibrating solutions of that concentration. The ALJ summarized Lorimer's testimony as follows:
The Casco standards are the focus of much of the expert debate which is at the heart of much of this case. The standards are solutions, purportedly containing 10, 20, 30, and 40 mmol/l of tCO2. According to Lorimer, they are sampled in the machine and the operator uses the readings in order to determine whether the readings obtained from the sampling of these solutions produce a linear readout. If the resulting numbers produce a straight line, within a degree of tolerance, then the calibration of the instrument established by use of the Beckman standards is verified. If the line becomes curvilinear the machine is not properly calibrated and must be adjusted.
The instructions accompanying the Casco solutions state that those solutions are intended for use only as checks on linearity to confirm calibration readings. Lorimer also testified that the ion-specific-electrode method used by the Commission has gained general scientific acceptance, and other jurisdictions have adopted it. (In response to this Court's request for supplemental briefs, the Attorney General confirmed that “the Illinois Racing Board Lab, the Maryland Racing Commission Lab, and the Michigan Department of Agriculture Lab all use the Commission's technique [.]”) New Jersey began using the method in March 1997.
Respondent's expert, Dr. Jonathan Foreman, an Associate Professor of Equine Internal Medicine at the University of Illinois, presented contrary testimony. He stated that calibrating the Beckman instrument with the Beckman solutions cannot “necessarily verify that the calibration is correct.” Dr. Foreman testified that the Beckman instrument had to be calibrated with reference to a 40 mmol/l standard solution (a solution not supplied by Beckman) to measure correctly in the 30 to 40 mmol/l range, the relevant range of interest for equine blood testing as set forth in the Commission's regulations.
In respect of the specific test performed on Ramses, Dr. Foreman expressed the view that the results from those tests over-reported the tCO2 concentration by about 1 mmol/l. He testified that if the readings on the confirmatory test had been adjusted based on a calibration using the Casco standards, the tCO2 value would have been 36.3 or 36.4 mmol/l, within the permissible range.
In rebuttal, Lorimer, the Commission's expert, disagreed. He testified that the Casco solutions have no carbon dioxide traceability and, therefore, the tCO2 concentrations can only be estimated. He stated that using the Casco solutions to calibrate the Beckman instrument “does not adhere to the more rigorous procedures used in analytical chemistry and forensic science,” and that no authoritative body of science has established or recommended the Casco standards for calibration. In support of his conclusion, Lorimer cited a letter that he had received from a Casco official, Dr. John Nagle, confirming that the Casco solutions were intended to establish the linearity of the measurements, not to calibrate the instrument. Wrote Dr. Nagle: “The [Casco] product is intended to measure the overall linearity of the analytical method employed for the [carbon dioxide] measurement; it is not a calibrator.”
The Commission also presented a report and testimony from Dr. Kenneth McKeever, an equine exercise physiologist and professor at Rutgers University. He testified that he had found the calibration of the instrument by the Commission's laboratory to be “impeccable.” Dr. McKeever explained that the laboratory's results were reliable because they were based on calibrations using the Beckman standards, which are certified traceable standards. (To establish certified standards, a company must comply with rules established by the National Institute of Standards and Technology.) Dr. McKeever stated that the Casco standards are not mixed with the same degree of accuracy or certainty as the Beckman standards, and thus are not suited to calibrating the instrument. Dr. McKeever testified that it was not necessary to calibrate the Beckman instrument at the top and bottom of the range of interest, so long as the machine's linearity was reliably established. He concluded that the Commission's laboratory results were “highly repeatable [and] accurate.”
In addition to the testimony of those three experts, the ALJ considered numerous exhibits and documentary evidence. Based on that record, the ALJ concluded that the Commission had established that Ramses, with its level of tCO2 above the concentration of 37 mmol/l, was in violation of N.J.A.C. 13:71-23.3A(a), and that the level of concentration did not reflect the horse's normal physiology. More specifically, the ALJ stated:
Based upon the record as it stands, the Commission has demonstrated that the use of the Beckman instrument is a scientifically accepted method for analyzing [tCO2] concentrations in blood plasma. In addition, the greater weight of the scientific testimony and evidence presented establishes that the use of the Beckman standards for calibrating the Beckman instrument is the generally accepted procedure, and that the use of Casco standards to demonstrate linearity is within the intended use advised by the manufacturer and also generally accepted. The record does not support any conclusion that the use of the Casco standards as primary calibrators is generally accepted within the scientific community, or that the scientific community rejects the calibration of the Beckman instrument with Beckman standards even where the value for the analyte [is] expected to be in excess of 30 mmol/l.
The Commission adopted the ALJ's decision as its final decision in May 1999, and stayed the imposition of penalties until the case could be appealed to the Appellate Division.
In an unreported opinion, the Appellate Division reversed. In essence, the panel concluded that it did not have to decide which solutions, the Beckman or the Casco, should be used to calibrate the instrument. Instead, the court stated that “the critical question in this case is whether the Beckman instrument as used by the Commission can measure the tCO2 in equine blood plasma with the precision required for enforcement of the 37 mmol/l regulatory standard.” In addressing that question, the panel concluded that “the evidence does not show that there is general acceptance within the scientific community for use of the Beckman analyzer in accordance with the Commission laboratory's procedures to determine concentrations of tCO2 with the precision required by the wording of N.J.A.C. 13:71-23.3A.”
The court also cited the results of testing performed on another horse that measured tCO2 levels at 36 mmol/l in an initial screening test and 38.3 mmol/l in a confirmatory test. The panel concluded overall that “the test results are subject to random errors of magnitudes which are significant for regulatory purposes[.]” Accordingly, the court held that the Commission had failed to carry its burden of demonstrating that the level of tCO2 in Ramses exceeded the regulatory limit.
The legal tenets governing our review may be stated briefly. Generally, an appellate court does not substitute its judgment of the facts for that of an administrative agency. Clowes v. Terminix Int'l, Inc., 109 N.J. 575, 587, 538 A.2d 794 (1988). In reviewing administrative adjudications, an appellate court must undertake a “careful and principled consideration of the agency record and findings.” Riverside Gen. v. N.J. Hosp. Rate Setting Comm'n, 98 N.J. 458, 468, 487 A.2d 714 (1985). “If the Appellate Division is satisfied after its review that the evidence and the inferences to be drawn therefrom support the agency head's decision, then it must affirm even if the court feels that it would have reached a different result itself.” Clowes, supra, 109 N.J. at 588, 538 A.2d 794.
Stated differently, if the agency's finding “ ‘is clearly a mistaken one and so plainly unwarranted that the interests of justice demand intervention and correction, then, and only then, [the appellate court] should appraise the record as if it were deciding the matter at inception and make its own findings and conclusions.’ ” Ibid. (quoting State v. Johnson, 42 N.J. 146, 162, 199 A.2d 809 (1964)). In the same vein, the choice of accepting or rejecting testimony from witnesses resides with the administrative agency, and so long as that choice is reasonably made it is accorded deference on appeal. Renan Realty Corp. v. State Dep't of Cmty. Affairs, Bureau of Hous. Inspection, 182 N.J.Super. 415, 421, 442 A.2d 614 (App.Div.1981).
Lastly, courts grant deference to agency expertise on technical matters, “where such expertise is a pertinent factor.” Close v. Kordulak Bros., 44 N.J. 589, 599, 210 A.2d 753 (1965). When resolution of a legal question turns on factual issues within the special province of an administrative agency, those mixed questions of law and fact are to be resolved based on the agency's fact finding. Boss v. Rockland Elec. Co., 95 N.J. 33, 42, 468 A.2d 1055 (1983).
Based on those tenets, we are satisfied that the Appellate Division erred in concluding that the Commission had not met its burden of demonstrating the reliability of the Beckman test generally and as applied in this instance. In these circumstances, we accord due deference to the Commission because its laboratory procedures involve technical matters falling within the special province of the agency's expertise. Relying on that expertise, the Commission adopted the ALJ's disposition. In so doing, the agency considered evidence from at least two experts supporting the reasonableness of its choice of method for demonstrating the precision of the Beckman instrument. The ALJ himself appeared to be persuaded particularly by Dr. McKeever's testimony that the use of the Beckman solutions to test for the accuracy of the instrument was appropriate.
Respondent's expert suggested that different solutions, the Casco solutions, should have been used to verify the precision of the instrument. The Commission's expert, Lorimer, contested that position using support from Casco's own company representative, Dr. Nagle. Similarly, the Commission's other expert, Dr. McKeever, disputed the explanation proffered by respondent's expert on the use of the Casco standards.
In sum, two qualified experts and documentary evidence supported the Commission's laboratory procedures. Accordingly, we hold that there is ample evidence in this record to support the Commission's determination that the testing of Ramses for tCO2 yielded a valid measurement in excess of the regulatory limit. Under the standards of appellate review that apply in this setting, the Commission's determination must be upheld.
The judgment of the Appellate Division is reversed. The matter is remanded to the Commission for reinstatement of its final decision.
For reversal-Chief Justice PORITZ and Justices COLEMAN, LONG, VERNIERO, LaVECCHIA and ZAZZALI-6. Opposed-none.