*780Sometimes there's a mismatch between what's at stake in a case before us and how interesting the legal issue involved is likely to be to the lay reader. This case is a prime example.
The dispute is an important one. Gary Woessner lost his life *781after an accident at work. His widow, Carmen, seeks death benefits under the Kansas Workers Compensation Act.
But the Kansas Legislature has provided that the employer isn't liable to pay benefits under that Act if an employee was impaired by drugs at the time of a work accident and that impairment contributed to the accident. Here, there's a drug test showing that Gary had marijuana metabolites in his system when he fell 15 feet from a jobsite catwalk for no apparent reason. If he was impaired by marijuana and that contributed to his fall, the employer has no liability for his death.
So whether Gary was impaired is an important issue. And the employer's liability-or lack of it-mainly hinges on that drug-test result.
That's where the legal issue before us comes into focus. For every piece of evidence *995in either an administrative hearing (which is where the evidence is presented in workers'-compensation cases) or in court, the party offering the evidence must present some foundation for its admission. By foundation, we mean a showing that the evidence is reliable enough that we should consider it along with other evidence in the case. And that's the issue we have to focus on here: was the foundation the employer presented for the admission of this drug-test result sufficient?
Unfortunately, at least for the lay reader, to answer that question we must wade through a series of provisions in the Workers Compensation Act, along with one regulation adopted administratively to implement the Act and several Kansas court decisions in workers'-compensation cases going back many decades. Here's a summary of what we found from doing that work.
First, the general rule for administrative proceedings, including workers'-compensation hearings, is that the rules for the admission of evidence aren't as strict as those used in court proceedings. Things are more informal, and the parties can present their cases without dotting every i and crossing every t-things that are expensive to do. Under these rules, hearsay evidence (evidence about what someone who isn't testifying as a witness said or reported) can be admitted as long as we're reasonably confident it's reliable.
Second, under that general rule, Gary's drug-test result should *782be admissible if a reasonable foundation is shown that they're reliable, even if testimony isn't offered from every person involved in the testing and chain of custody. Gary's employer presented ample evidence about the reliability of the drug test. So the test result should be admissible unless there's some specific provision in the Workers Compensation Act that might preclude it.
Third, we find no statutory provision in the Act (and no regulation adopted to implement the Act) that would make it harder to get the drug-test result in Gary's case admitted. In that conclusion, we disagree with the Workers Compensation Appeals Board, which concluded that the drug-test result was inadmissible under a provision of the Act and an administrative regulation.
Having concluded that the Board incorrectly excluded the drug-test result, we must explain that there are two steps to determining whether the employer's liability is eliminated. Step one is the drug-test result. When, as here, it shows that the employee had marijuana in his system at a sufficient level, there's a conclusive presumption under the Act that the employee was impaired. Step two, though, is determining whether that impairment caused or contributed to the accidental injury. In that step, the employee has the burden to show by clear and convincing evidence that the impairment didn't contribute to the injury.
Although the Board first concluded-in error-that the drug-test result should not be considered at all, the Board separately concluded as an alternative basis for its decision that Carmen had shown that Gary's impairment didn't contribute to the accident. But the Board's alternative finding was made in a single paragraph of only four sentences; the Board did not explain how it found the testimony of one lay witness so persuasive on this point. Since the Board's decision focused mostly on the admissibility of the drug-test result and its explanation of this alternative conclusion was so cursory, we are not satisfied that the Board carefully weighed the evidence on this point. We therefore return the case to the Board for further consideration of this step in the analysis; the Board should consider the drug-test result and all the other evidence.
With that overview, we will proceed with an opinion that more fully explores these issues. Since the foundation for the drug-test *783results is the key to this appeal, we must first set out in detail the evidence presented for that purpose. Once we have set out that evidence, we will explain why we have concluded that the employer presented a sufficient foundation to have the drug-test result admitted and considered along with other evidence in the case.
FACTUAL AND PROCEDURAL BACKGROUND
Before we go through the evidence related to the drug test, we must explain how disputes *996in a workers'-compensation claim are resolved. Although there can be preliminary hearings if there are disputes about temporary payments or medical treatment, most issues get decided at what's called the "regular hearing." That's the equivalent of a court trial, though the rules are a bit different, as we'll explain more fully later in the opinion.
But there's one difference we should note up front. Unlike a court trial, in which all the evidence is presented in front of a judge-who can then rule on objections as they are made-most of the evidence in a workers'-compensation case comes through depositions. At those depositions, a witness is present, along with attorneys for the parties and a court reporter, and the witness gives sworn testimony, just as if in court. But the administrative law judge isn't present for the depositions, so the attorneys make objections that must be ruled on later. That leaves both parties guessing a bit about how to proceed.
Another aspect of the way workers'-compensation cases are handled makes this complicated for the attorneys and the parties-the administrative law judge doesn't have the final say on what evidence is admitted. That's because either party may appeal to the Workers Compensation Appeals Board, and the Board makes its own independent judgment. So even when the administrative law judge rules that something is admissible at the regular hearing, there's still the possibility that the Board might take a different view.
With that overview, let's review the evidence presented about Gary's drug testing. Through several exhibits and the testimony of witnesses, we'll learn that a urine drug screen run at Stormont Vail Hospital in Topeka when Gary was being treated on an emergency basis tested positive for THC, marijuana's psychoactive ingredient.
*784We'll also learn that a part of the sample taken when Gary was treated at Stormont Vail was later sent to LabCorp, an independent testing laboratory; LabCorp reported the presence of metabolites of marijuana. (These marijuana metabolites are the inactive compounds that remain after the body has metabolized THC.) Of course, we must focus not only on these reported results but also on the foundation provided to show that the results are reliable enough to consider in a case about workers'-compensation benefits.
The Regular Hearing
The parties agreed at the start of the regular hearing that the only contested issue was whether marijuana intoxication relieved the employer from liability to pay workers'-compensation benefits. Over objections from Carmen's attorney, the administrative law judge admitted two exhibits about the drug tests: Exhibit B, an affidavit and business records from Shelley D'Attilio, the lab director at Stormont Vail; and Exhibit C, an affidavit and business records from David St. John, the lab supervisor for the LabCorp facility in Mississippi where Gary's urine sample was tested.
Carmen's attorney objected to the D'Attilio affidavit on two grounds: hearsay and foundation. The administrative law judge overruled the objections, noting that "hearsay is technically admissible at administrative proceedings" and that the admission of Exhibit B would not prevent Carmen from challenging the statements made in the affidavit.
Carmen's attorney made the same hearsay and foundation objections to the St. John affidavit. The judge said that the LabCorp material "goes to testing, ... not medical treatment" and that no statute precluded the admission of hearsay evidence that wasn't related to medical treatment. The judge again noted that Carmen could still challenge the accuracy of the information with argument and other evidence.
D'Attilio's affidavit, labelled a chain-of-custody affidavit, told what had happened to Gary's urine specimen. She said that she knew about these facts based on her review of records kept by Stormont Vail in the ordinary course of business and her knowledge of *785the procedures of the lab she directs. D'Attilio identified each step in handling the specimen:
• Hospital employee Sarah Persons took the sample for medical testing at about 5:41 p.m. on December 19, 2014.
*997• About 10 minutes later, laboratory employee Kerry Coulter received the specimen.
• Another lab employee, Larry Reid, then put a portion of the specimen on an analyzer for testing; the machine showed a positive result for THC.
• Yet another lab employee, Deb Redler, verified the positive result at 6:19 p.m.
• Under standard procedures, the specimen then would have been transferred to a locked toxicology refrigerator for the night; the next morning, it would have been moved to a freezer for long-term storage. Only Stormont Vail lab staff doing toxicology work had access to that freezer.
• After getting a request for further testing, lab employee Paul Brackey retrieved the sample from the freezer, thawed it, and sent a portion to LabCorp in a sealed container. That container, accompanied by a chain-of-custody form, was given to a dedicated courier who took samples daily to LabCorp. The remaining sample was then returned to long-term storage in the Stormont Vail toxicology freezer.
Two documents were attached to the D'Attilio affidavit. One was the chain-of-custody form that was apparently sent to LabCorp. At the top, it referred to the specimen by number-0089146682. That form was signed by Paul Brackey and said that the specimen had been released to the "LabCorp Courier." The other document was the report of the drug-test result done at Stormont Vail. It showed a positive result for THC with a cutoff value of 50 nanograms per milliliter. The report also said: "Results not confirmed. May not meet forensic requirements. Confirmation by GC/MS available upon request."
In his affidavit, St. John, the lab supervisor for LabCorp's facility in Southaven, Mississippi, said he had reviewed the business records from the lab related to Gary's testing and that these *786records had been made at or near the time of each event noted. He said LabCorp received sample No. 0089146682 "with seals intact and with no evidence of tampering" and that lab personnel then handled the specimen under proper procedures, including keeping chain-of-custody documentation. He said that lab staff used the gas chromatography /mass spectrometry (GC/MS) method, "which provides accurate identification and quantitation of a drug or drug metabolite." He said that the testing found a positive result for marijuana metabolites, a result confirmed and reported by a certifying scientist. St. John also said that LabCorp's Southaven lab is federally certified for drug testing.
Several documents were attached to St. John's affidavit:
• His letter report stating that the test result showed marijuana metabolites at a level of 189 nanograms per milliliter.
• A subpoena for the "Litigation Claim Packet" related to Woessner's test.
• A "Laboratory Data Package" that contained the lab's documents related to testing, confirmation, and chain-of-custody for Woessner's sample. The package also had a curriculum vitae for Lance Presley, the lab director; St. John, the lab supervisor; and Kathryn Atkins, the scientist who confirmed the GC/MS result for Woessner.
Deposition of Dr. Nason Lui
Dr. Nason Lui was the surgeon on duty in the Stormont Vail emergency room when Gary was transferred there from a smaller hospital after his accident. Dr. Lui said Gary was "basically unconscious" when he arrived, and his status triggered staff in the emergency room to perform a set of tests, including a drug screen. Dr. Lui's name was listed as the ordering physician on Stormont Vail's records, but he said that because things must move quickly in the emergency room, tests like Gary's drug screen are automatically ordered based on hospital protocols and the patient's status.
Even though the test was ordered automatically, Dr. Lui said that it was ordered to give Gary good medical treatment; especially for an unconscious patient, treating doctors need lab tests to check *787for possible issues. Dr. Lui noted the positive THC test in his hospital chart note. *998The hospital's toxicology report showing the positive THC test was an exhibit at the deposition. Dr. Lui said that this document was a hospital record that had been kept in the ordinary course of treating Gary. The report noted that Gary's urine had been obtained by a catheter and showed a positive result for THC with the 50 nanogram-per-milliliter cutoff level.
Deposition of Shelley D'Attilio
D'Attilio is the lab director at Stormont Vail Hospital. She said she was asked by a hospital attorney to look into the procedures involved with Gary's urine sample and its testing. She mostly reviewed the lab's records, though she also talked with Paul Brackey, the lab scientist in the hospital's toxicology group who forwarded a portion of the sample to LabCorp.
She also explained a difference in the chain-of-custody procedures used when the testing is being done for medical purposes rather than to meet some legal requirement (like the confirmatory test obtained here from LabCorp). When the testing is for medical purposes, she said that strict chain-of-custody documentation rules don't apply. For that reason, while the records show that the urine sample was obtained by Sarah Persons and was received in the lab by Kerry Coulter, she said that someone would send it through the hospital's pneumatic tube system to the lab. We don't necessarily know who did that. Nor did D'Attilio report specifically who placed the sample into the locked toxicology refrigerator; she merely reported that this was the normal practice in the lab.
D'Attilio said that all the records she reviewed had been maintained by Stormont Vail Hospital in the ordinary course of business; as lab director, she was responsible for maintaining these records. She based the statements in her affidavit on her review of these lab records. Her review of the lab records about Gary's drug test showed nothing suggesting any irregularity or error. As for the test result from her hospital's lab (a positive indicator for THC), she said that "[n]othing in my investigation would indicate to me that there was a possibility that this result was incorrect." From her *788review of the records, lab employee Larry Reid put the sample into a machine to be tested, the machine produced a resulting value, and lab employee Deb Redler reviewed and "verified" that result. When she did so, the test result would have been transferred to Gary's hospital record so that treating physicians would have had it available to them.
D'Attilio also said that proper procedures were followed to prevent contamination of Gary's sample while it was kept at Stormont Vail. She said that LabCorp regularly does follow-up testing for Stormont Vail; the hospital sends samples to LabCorp's Kansas City facility through a daily courier.
Deposition of Dr. Christopher Long
Dr. Christopher Long has a Ph.D. in toxicology and is a board-certified forensic toxicologist. He was hired as an expert witness by Labor Max and its insurance carrier. Dr. Long directs the forensic toxicology laboratory at the St. Louis University School of Medicine, has directed federally certified labs that do drug testing, and has done national inspections of federally certified labs for federal authorities.
He reviewed the lab tests and documentation from both Stormont Vail and LabCorp. That included Exhibit C, the affidavit from LabCorp lab supervisor St. John and the LabCorp testing records. Part of Exhibit C was called the LabCorp litigation package; Dr. Long said that he could determine whether appropriate chain-of-custody procedures were followed from the litigation package. Dr. Long said that he had inspected "most, if not all," of the LabCorp labs around the country, apparently during the time he was doing inspections of federally certified labs.
Dr. Long explained the differences between the test done at Stormont Vail, called an immunoassay test, and the one done at LabCorp using gas chromatography /mass spectroscopy, also called GC/MS. An immunoassay test measures an allergic reaction in the test tube, but it can have false-positive results and must be confirmed. A GC/MS test is "the gold standard in the field" and *999identifies the molecular structure of the material being tested.
Dr. Long said that the initial test at Stormont Vail had a cutoff *789level of 50 nanograms per milliliter; if the result doesn't exceed 50 in that test, it's reported as negative. He said that the Stormont Vail test result showed a level above 50 nanograms per milliliter. Dr. Long testified that the LabCorp GC/MS testing showed a marijuana-metabolite level of 189 nanograms per milliliter, which confirmed the earlier test result. Dr. Long said that he could rule out the possibility that Gary's test was a false positive.
The LabCorp report had a comment: "Temperature out of acceptable range." Dr. Long explained that labs must check the temperature of urine samples as a guard against cheating the test-a person who comes to get drug tested but has a container of drug-free urine with them. In that case, the urine sample would be closer to room temperature than body temperature, revealing an attempt to cheat. Here, though, the sample had been taken by catheter months before and then frozen, so the temperature wouldn't be expected to be in the acceptable range under normal urine-testing protocols. Dr. Long said that the temperature of the sample would have had no impact on the test result here.
Based on the level of metabolites in the LabCorp test, Dr. Long said that Gary had ingested marijuana "within 6 to 12 hours, maybe 24 hours on the outside." He said that "the concentration in the urine reflects the acuteness of the dose." Because the LabCorp testing looked for marijuana metabolites (not the active and hallucinogenic ingredient, THC, itself), Dr. Long said that he could not give an opinion about Gary's level of impairment at the time of his accident.
Dr. Long said that he had gone through all the lab records "in detail." His conclusion: "There's no doubt in my mind that the tests accurately reflect Mr. Woessner's specimen." He added that the opinions he had given in his testimony and written report were to a reasonable degree of scientific certainty in the field of toxicology.
Decision of the Administrative Law Judge
The administrative law judge determined that the drug-test results, including LabCorp's GC/MS test, were admissible. That triggered the statutory presumption that Gary was impaired at the time *790of his accident, and the judge found that Carmen hadn't presented sufficient evidence to rebut that presumption. Based on these conclusions, the judge ruled in favor of Labor Max and its insurance carrier and did not order any further payments to Carmen. Carmen then appealed to the Workers Compensation Appeals Board.
Decision of the Workers Compensation Appeals Board
The Board concluded that the LabCorp GC/MS test result was not admissible. The Board relied on a statute, K.S.A. 2014 Supp. 44-501(b)(3), and a regulation, K.A.R. 51-3-5a. The statute provides several requirements for the admission of drug-test results from "a sample collected by an employer," including that "the foundation evidence ... establish, beyond a reasonable doubt , that the test results were from the sample taken from the employee." (Emphasis added.) K.S.A. 2017 Supp. 44-501(b)(3). Although this sample was collected by Stormont Vail, not Gary's employer, the Board apparently applied this statutory provision because Labor Max had requested the additional LabCorp test. The Board noted that no one who directly handled Gary's sample had testified and concluded: "Without a demonstrable and reliable chain of custody, it would be impossible to conclude, beyond a reasonable doubt, that a given sample was from [Gary] and that the same sample was used for the drug test."
The regulation cited by the Board, K.A.R. 51-3-5a, requires that some "reports, records, or statements" not be admitted without the testimony of the person who made them. We'll discuss that regulation in greater detail later in this opinion, but the Board applied it to the LabCorp test result. Because no LabCorp personnel testified at the regular hearing or in a deposition, the Board held that the LabCorp test result couldn't be admitted.
Without that test result, there would be no statutory presumption that Gary was impaired *1000at the time of his accident. The Board said the other evidence of impairment wasn't sufficient to show that any impairment from marijuana usage contributed to the accident, so Carmen was entitled to the remaining benefits due under the Workers Compensation Act.
In the alternative, the Board said that "even if the [LabCorp] ...
*791test results are admissible, there was clear and convincing evidence rebutting the presumption that [Gary] was impaired at the time of his accident." The Board cited the testimony of one of Gary's coworkers, Thomas McGraw, who said that he hadn't noticed any signs that Gary was impaired before the accident.
ANALYSIS
The key question is whether the LabCorp test result should have been admitted into evidence and considered by the Board. Before we get to that, though, we need to put the general rules for the fact-finding process in workers'-compensation proceedings in context.
The Rules of Evidence for Workers'-Compensation Hearings
The evidence is presented in an administrative proceeding, not a court trial, and that makes a difference. In court, formal rules of evidence apply-in Kansas, they are set out in statutes called the Kansas Rules of Evidence, found at K.S.A. 60-401 to 60-485. But those rules don't generally apply in administrative hearings, and the Legislature has specifically made them inapplicable to workers'-compensation hearings.
A provision in the Workers Compensation Act, K.S.A. 2017 Supp. 44-523(a), says that neither the administrative law judge nor the Board are "bound by technical rules of procedure." And by " 'technical rules of procedure,' " the Act refers to the Kansas Rules of Civil Procedure and "particularly" the Kansas Rules of Evidence. Roberts v. J.C. Penney Co. , 263 Kan. 270, 278, 949 P.2d 613 (1997). That means that the Kansas Rules of Evidence don't apply, so the admissibility of evidence is "more liberal in compensation cases" than in court cases. Gasswint v. Superior Industries Int'l-Kansas, Inc. , 39 Kan. App. 2d 553, 560, 185 P.3d 284 (2008).
One of the main differences is the treatment of hearsay evidence-an out-of-court statement made by someone who isn't testifying when presented to prove that the out-of-court statement was true. That's part of what we're dealing with here-the out-of-court statement from David St. John, the LabCorp lab supervisor, about the test result. Labor Max wants us to accept that test result as true. And since St. John just sent an affidavit and wasn't deposed, *792that makes his statements hearsay when offered to prove that what he says is true.
But hearsay evidence can be admitted in administrative proceedings-and for good reason. It costs money to bring everyone involved in every aspect of a dispute either to the hearing or to a deposition. The time of the attorneys and the transcripts of the court reporters can be expensive, and the time it takes to assemble all of these people can delay getting a result. So administrative proceedings are streamlined to strike a balance between strictly following all the evidentiary rules and allowing just about anything to be considered. The goal for these proceedings is to get a fair result within a reasonable time and at a reasonable cost. To help accomplish that goal, for most Kansas administrative hearings "[e]vidence need not be excluded solely because it is hearsay." K.S.A. 77-524(a).
That specific provision, part of the Kansas Administrative Procedure Act, doesn't apply to workers'-compensation proceedings. But K.S.A. 77-524(a), based on a model state-administrative-procedure statute, is quite similar to K.S.A. 2017 Supp. 44-523(a), the provision we've mentioned from the Workers Compensation Act. Both provide that officers deciding what comes into evidence aren't "bound by technical rules of procedure" and "shall give the parties reasonable opportunity to be heard and to present evidence." K.S.A. 2017 Supp. 44-523(a) ; K.S.A. 77-524(a). And while there's no statutory statement in K.S.A. 2017 Supp. 44-523(a) indicating that this means that hearsay evidence is allowed, that has long been the understanding under the Workers Compensation Act. See *1001Neal v. Hy-Vee, Inc. , 277 Kan. 1, 22, 81 P.3d 425 (2003) (finding hearsay evidence sufficient to support initial award); Shindhelm v. Razook , 190 Kan. 80, 84, 372 P.2d 278 (1962) (finding no error in consideration of hearsay evidence in workers'-compensation death case); Pence v. Centex Construction Co. , 189 Kan. 718, 371 P.2d 100 (1962) (citing cases from 1931 to 1961 allowing hearsay evidence in Kansas workers'-compensation hearings). Consistent with that understanding, an administrative regulation for workers'-compensation hearings explicitly says that "[h]earsay evidence may be admissible unless irrelevant or redundant." K.A.R. 51-3-8(c). *793Even so, that doesn't mean that all hearsay evidence must be admitted. When you get down to "Joe said that Mary said that Jim said that Bob said that he had heard that someone pushed Gary off the catwalk," we all get the sense that what we're hearing just isn't reliable enough to make the basis of a good decision. So what an administrative law judge or the Board must consider is whether the hearsay evidence is reliable enough to consider as part of the decision at hand. And that decision is made against the background considerations of K.S.A. 2017 Supp. 44-523(a) and K.A.R. 51-3-8(c) -that technical rules of evidence don't apply, evidence is more easily admitted than in civil and criminal trials, and evidence isn't rejected just because it's hearsay.
The Role of the Board and of the Appellate Court on Review
Decisions about the admission of evidence are made in workers'-compensation hearings by the administrative law judge or the Board, with the Board having final say. See K.S.A. 2017 Supp. 44-523(a) ; K.S.A. 2017 Supp. 44-555c(a). Here, we are considering whether the Board's decision to exclude hearsay evidence about the LabCorp test result was proper. Everyone recognizes that the test result is relevant-the dispute is about whether foundation evidence about the chain of custody was enough to admit evidence of the test result.
Evidentiary foundations come in all shapes and sizes, so there's no one-size-fits-all rule beyond the idea that the evidence be reliable enough to consider. In these situations, the party making the admissibility decision-here, the Board-has discretion to determine how much of a foundation is enough.
On appeal, we review that discretionary call for abuse of discretion. See State v. Lowery , 308 Kan. 1183, 1232, 427 P.3d 865 (2018) ; Garetson Brothers v. American Warrior, Inc. , 51 Kan. App. 2d 370, 383, 347 P.3d 687 (2015) ; 101 C.J.S. Workers' Compensation § 1462. That means that we reverse the Board's decision only if that decision was based on a factual or legal error or if no reasonable person would agree with the decision. See Garetson Brothers , 51 Kan. App. 2d at 383, 347 P.3d 687 ; Rogers v. ALT-A&M JV , 52 Kan. App. 2d 213, 217, 364 P.3d 1206 (2015).
*794With these standards in mind, we turn to the Board's decision. It decided that a statute, K.S.A. 2017 Supp. 44-501(b)(3), and an administrative regulation, K.A.R. 51-3-5a, precluded the admission of the LabCorp test result. As we'll explain, though, the Board's analysis of these provisions is flawed, and a decision based on a legal error is an abuse of discretion. Garetson Brothers , 51 Kan. App. 2d at 383, 347 P.3d 687.
The Statutory Provision Relied on by the Board to Exclude the Lab Result
Let's start with the statutory provision, K.S.A. 2017 Supp. 44-501(b)(3). The full statutory section, K.S.A. 2017 Supp. 44-501, provides that employees don't get compensated for an injury in some situations, such as when the employee deliberately fails to use protection equipment furnished by the employer. See K.S.A. 2017 Supp. 44-501(a)(1)(C). In our case, the focus is on subsection (b), which disallows compensation when the employee's injury, disability, or death was contributed to by the use or consumption of alcohol or drugs. See K.S.A. 2017 Supp. 44-501(b)(1)(A).
If there's a lab test confirmed through GC/MS testing showing a concentration of a listed drug above certain levels, then the employee is "conclusively presumed ... impaired,"
*1002and there's a rebuttable presumption that the impairment contributed to the employee's injury, disability, or death. See K.S.A. 2017 Supp. 44-501(b)(1)(C) ; K.S.A. 2017 Supp. 44-501(b)(1)(D). For marijuana metabolites, a confirmed test result at or above 15 nanograms per milliliter triggers the impairment presumption; Gary's level of 189 far exceeds that. But the employee may overcome that presumption by presenting "clear and convincing evidence" that the impairment didn't contribute to the injury, disability, or death. K.S.A. 2017 Supp. 44-501(b)(1)(D).
In addition to these substantive rules-under which an employee may lose workers'-compensation benefits-the Kansas Legislature also provided some rules about the admissibility or inadmissibility of test results. In K.S.A. 2017 Supp. 44-501(b)(2), the Legislature set out a rule of admissibility , providing that test results "shall be admissible" if the employer establishes one of several things. For *795example, if the employer shows that a test result was obtained during an autopsy and not at the employer's direction, the result is admissible. K.S.A. 2017 Supp. 44-501(b)(2)(B). But none of the specific situations mentioned in subsection (b)(2) match the facts of our case. Neither party suggests subsection (b)(2) applies here, and the Board did not rely on it.
The Board instead relied on subsection (b)(3), which sets out a rule of inadmissibility , making test results inadmissible in some circumstances. K.S.A. 2017 Supp. 44-501(b)(3) applies only to "the results of a chemical test performed on a sample collected by an employer ." (Emphasis added.) For those samples, the employer must show several specific things, such as that the sample was collected promptly and tested by an approved lab. But there's also an evidentiary foundation requirement for these test results: "[T]he foundation evidence must establish, beyond a reasonable doubt, that the test results were from the sample taken from the employee." K.S.A. 2017 Supp. 44-501(b)(3)(E). Relying on this statutory provision, the Board said that without testimony from someone who had directly handled Gary's sample, "it would be impossible to conclude, beyond a reasonable doubt, that a given sample was from [Gary] ...."
But subsection (b)(3) doesn't apply here at all. It applies only to samples "collected by an employer," and the sample here was taken by Stormont Vail Hospital personnel. The Board made a legal error by relying on this statute to exclude the lab-test result.
The Administrative Regulation Relied on by the Board to Exclude the Lab Result
The Board also cited an administrative regulation, K.A.R. 51-3-5a. It's headed "Procedure for preliminary hearings," the proceedings that take place early in a case to determine things like whether the employee gets temporary benefits or some specific medical care. Since the lab evidence was presented at the regular hearing-not a preliminary one-you'd think that this regulation would be unlikely to apply. But the Board seized on one sentence in the regulation that can apply in some circumstances when a final *796award is made. We've set out K.A.R. 51-3-5a in full below with that sentence italicized:
"Procedure for preliminary Hearings . (a) Medical reports or any other records or statements shall be considered by the administrative law judge at the preliminary hearing. However, the reports shall not be considered as evidence when the administrative law judge makes a final award in the case, unless all parties stipulate to the reports, records, or statements or unless the report, record, or statement is later supported by the testimony of the physician, surgeon, or other person making the report, record, or statement. If medical reports are not available or have not been produced before the preliminary hearing, either party shall be entitled to an ex parte order for production of the reports upon motion to the administrative law judge.
"(b) If the decision of the administrative law judge is not rendered within five days of the hearing, the applicant's attorney shall notify the director, who shall make demand upon the administrative law judge for this decision.
"(c) In no case shall an application for preliminary hearing be entertained by the *1003administrative law judge when written notice has not been given to the adverse party pursuant to K.S.A. 44-534a." (Emphasis added.)
With that text before us, let's break down what K.A.R. 51-3-5a does. It's clear that it is concerned mainly with preliminary hearings, not the later hearing at which evidence is presented to determine a final award. All three of the subsections relate to preliminary hearings.
The first sentence allows "[m]edical reports or any other records or statements" to be considered at a preliminary hearing. That way, you don't have to bring doctors and other witnesses to a hearing on preliminary matters. The second sentence then provides that "the reports" won't be considered on a final award unless the parties stipulate to their admission or they are "supported by the testimony of the physician, surgeon, or other person making the report, record, or statement."
That second sentence tells us that "the reports shall not be considered as evidence" without testimonial support. But what does that encompass? In the first sentence (and the third one), there's a reference to "medical reports," so "the reports" certainly include the medical reports considered at a preliminary hearing. In our case, though, the LabCorp test results weren't submitted at a preliminary hearing; they were submitted only at the final hearing.
There is a bit of ambiguity in the second sentence, though.
*797Although the early reference to "the reports" seems to refer back to "[m]edical reports" from the first sentence, later portions of that second sentence refer to "the reports, records, or statements" being "supported by the testimony of the physician, surgeon, or other person making the report, record, or statement." So "the reports" may refer to reports, records, or statements. Even so, the inclusion within a regulation dealing only with preliminary hearings and the reference in the second sentence to "the reports" still suggest to us that we're talking only about things that were presented at a preliminary hearing. And the third sentence again refers to "medical reports," reinforcing the idea that we're talking about materials related to the employee's medical condition being used at the preliminary hearing.
Under this regulation, then, a party can get medical records, reports, or statements considered at a preliminary hearing without a witness-but what's admitted that way at the preliminary hearing doesn't come in at the final hearing without a witness. And this regulation's requirement that a report, record, or statement be "later supported by ... testimony" reasonably applies only to documents that were first presented at a preliminary hearing without the required testimony. That's why the second sentence of subsection (a) refers to "the reports" after the first sentence talked about allowing medical reports to be considered at the preliminary hearing. And it's why the second sentence refers to having the report be "later supported" by testimony-the report would have been presented "earlier" without it.
Our understanding is consistent with the statutory provision governing preliminary hearings, K.S.A. 2017 Supp. 44-534a. It says that the preliminary hearing "shall be summary in nature" and that the findings made at it "shall not be binding in a full hearing on the claim, but shall be subject to a full presentation of the facts." K.S.A. 2017 Supp. 44-534a(a)(2).
The Board reads "reports, records, or statements" to apply to virtually anything, whether a medical record or not-and whether first presented at a preliminary hearing or not. That takes K.A.R. 51-3-5a far beyond its apparent reach. As we read K.A.R. 51-3-5a, it relates to medical statements and records that are considered at *798the preliminary hearing without a doctor's testimony. Later though, at least in nondeath cases, each side has the opportunity to have a doctor conduct a medical examination of the employee so that disability ratings may be obtained. See K.S.A. 2017 Supp. 44-515 and 44-516. Because of the importance of those examinations and ratings, another statute, K.S.A. 44-519, provides that "no report of any examination ... by a health care provider ... shall be competent evidence ... unless supported by the testimony of such health care provider *1004...." Read in light of K.S.A. 44-519 (requiring a doctor's testimony to support an examination report) and K.S.A. 2017 Supp. 44-534a (providing that preliminary hearings be summary in nature), the regulation, K.A.R. 51-3-5a, makes sense. It allows a doctor's report to be admitted at a preliminary hearing without testimony but warns all parties that testimony will be needed later. Reading K.A.R. 51-3-5a as the Board does-to apply to virtually all documents or statements-would eviscerate the general rule that hearsay evidence is admissible in workers'-compensation hearings.
We'll conclude our discussion of K.A.R. 51-3-5a with one final point: It must be read in the context of the overall Workers Compensation Act. Regulations adopted by an administrative agency don't override statutes adopted by the Legislature. So the regulations must be consistent with the underlying statutes. Barbury v. Duckwall Alco Stores , 42 Kan. App. 2d 693, 694, 215 P.3d 643 (2009).
Under the Act, strict evidentiary rules don't apply and hearsay evidence is generally admissible. The Board's broad reading of K.A.R. 51-3-5a, even if textually permissible, would be unlikely when this context is considered. In situations like this involving federal legislation, the late Justice Antonin Scalia would say that Congress wouldn't hide an elephant in a mousehole. See Whitman v. American Trucking Assns,Inc. , 531 U.S. 457, 468, 121 S.Ct. 903, 149 L.Ed.2d 1 (2001). A textualist justice who relied on statutory language, not legislative history, Justice Scalia provided an important lesson about how statutes are put together: You don't hide a really important provision in a subsection of an otherwise unimportant part of a set of complicated and interrelated statutory provisions.
*799Under that view-even if K.A.R. 51-3-5a were part of the statute itself-it would be improbable that the Kansas Legislature would have wanted to make such a major change to the evidentiary rules in an ancillary provision aimed at governing preliminary hearings. But this isn't a statute. It's a mere administrative regulation, not something enacted by the Legislature. The meaning given to it by the Board goes several steps beyond this regulation's appropriate reach.
In sum, the statute for preliminary hearings provides that they be summary in nature; the regulation provides that medical reports, records, and statements may be admitted at the preliminary hearing without a witness; and both statute and regulation provide that there be a full presentation at the final hearing on matters that were heard in summary fashion at the preliminary hearing. Nothing in either the statute or regulation provides that reports of lab testing-not first presented in a preliminary hearing-are subject to any special evidentiary rules.
Reviewing the Board's Decision under the General Evidentiary Rules for Workers'-Compensation Hearings
Now that we've established that no specific statute or regulation precludes the admission of the LabCorp test result, we turn to the next part of the abuse-of-discretion analysis. At this step, we must ask whether the Board abused its discretion when it found the foundation for admission of the test result to be inadequate. We now look to see whether any reasonable person could have found this foundation inadequate while considering the appropriate legal standards. And under those standards, hearsay evidence is generally admissible, though the Board still must look to see whether the test result was reliable enough for it to consider. Considered under those standards, Labor Max presented ample evidence supporting the reliability of the test result.
The evidence about how the urine sample was obtained at Stormont Vail, how the hospital lab tested it, how the hospital stored the sample, and how the hospital then transferred the sample to LabCorp would meet evidentiary requirements under the Kansas Rules of Evidence. The hospital's lab director, Shelley D'Attilio, *800testified about each step of the process and the hospital lab's test result based on her review of business records kept under her supervision. So while she relied on some hearsay evidence-what was reflected in the hospital's records-there's *1005a hearsay-evidence exception in the Kansas Rules of Evidence allowing the admission of business records. See K.S.A. 2017 Supp. 60-460(m).
For the LabCorp part of the process, Labor Max submitted the affidavit of David St. John, the lab supervisor for the Mississippi testing facility that tested this sample, and testimony from Dr. Christopher Long, the toxicologist. Had St. John testified to the things in his affidavit, his testimony would have gone from hearsay status (an out-of-court statement) to admissible under the Kansas Rules of Evidence. He too explained that he had carefully reviewed business records kept in the ordinary course of business that would qualify, had he testified, for admission under K.S.A. 2017 Supp. 60-460(m).
But the submission of an affidavit from him instead of testimony doesn't automatically make his hearsay statements inadmissible. In workers'-compensation cases, "[h]earsay evidence may be admissible unless irrelevant or redundant." K.A.R. 51-3-8(c). This test result was neither irrelevant nor redundant.
It's significant that St. John works at the federally certified lab of a third party and has no interest in this case. And when the St. John affidavit and exhibits are combined with the testimony of Dr. Long, there's more than enough evidence to get the test result admitted. Dr. Long's academic qualifications haven't been challenged, and he fully reviewed all the test data from LabCorp. He said that he had determined from the records that proper chain-of-custody procedures were followed. He said he had "no doubt" that the test result was accurate.
The Board complained that no one who touched Gary's sample along the way testified. But even in court, lab results can be admitted with just the testimony of a business-records custodian. See, e.g., Shelton v. Consumer Products Safety Comm'n , 277 F.3d 998, 1010 (8th Cir. 2002) (approving the admission of lab test results as business records); Holbrook v. Precision Helicopters, Inc. , 162 Or. App. 538, 543-44, 986 P.2d 646 (1999) (same). And here the *801hearsay rule applicable to court proceedings-that hearsay evidence isn't admitted without an applicable exception, like the business-records exception-doesn't apply. Instead, hearsay evidence is admitted in workers'-compensation hearings as long as there are sufficient indicia of reliability. That standard is met here.
We've now established that the Board erred when it excluded the LabCorp test result. With the admission of that test result, the conclusive presumption of impairment under K.S.A. 2017 Supp. 44-501(b)(1)(C) applies and there's a rebuttable presumption that this impairment contributed to Gary's death. Given that presumption, Carmen can recover only if she shows by clear and convincing evidence that the impairment didn't contribute to the death. See K.S.A. 2017 Supp. 44-501(b)(1)(D). That means that she must show that it was highly probable that the impairment didn't contribute to Gary's death. See In re B.D.-Y. , 286 Kan. 686, Syl. ¶ 1, 187 P.3d 594 (2008).
The Alternative Basis for the Board's Decision
The Board did conclude, as an alternative basis for its ruling, that Carmen had met this standard. The Board relied on the testimony of Gary's coworker, Thomas McGraw. The Board said that McGraw had "observed [Gary] shortly before his accident and he did not notice [Gary] acting any differently than he normally did."
But the Board did not discuss Dr. Long's testimony that a layperson generally can't tell when someone has used marijuana. Dr. Long testified that alcohol impairment acts as a "depressant that's easy to see," but that "something [like marijuana] that causes alterations in your sensory input is very hard for a layperson to identify."
Given the very brief explanation by the Board of this alternative basis for its ruling and the high standard of proof applicable to it, we are not comfortable evaluating this alternative basis for the Board's decision without a more complete explanation of why it found McGraw's testimony to meet the clear-and-convincing evidence test given Dr. Long's related testimony. When an administrative agency's ruling isn't sufficient to allow for meaningful judicial review, the appellate *1006court may remand the case to the agency. Jones v. Kansas State University , 279 Kan. 128, 142, 106 P.3d 10 (2005). We *802conclude that the appropriate course here is to remand the case to the Board for further consideration. See K.S.A. 2017 Supp. 44-556(a) ; K.S.A. 77-622(b) ; Casco v. Armour Swift-Eckrich , 283 Kan. 508, 529, 154 P.3d 494 (2007) ; Jones , 279 Kan. at 142, 106 P.3d 10. On remand, the Board should fully consider the evidence, including the lab-test result, and is not bound by its previous ruling.
One Side Point
Before we conclude our opinion, we want to clear up one area of confusion in the parties' briefs. We haven't mentioned it until now because it doesn't affect our analysis, but Carmen made one additional argument that-had the factual premise for it been correct-would have been significant.
To this point in the opinion, we've simply referred to testing for marijuana metabolites. But the Kansas statute specifying that a conclusive presumption of impairment applies if the level of metabolites exceeds 15 nanograms per milliliter lists one specific marijuana metabolite in a footnote: Delta-9-tetrahydrocannabinol-9-carboxylic acid. See K.S.A. 2017 Supp. 44-501(b)(1)(C). Under that statute, the GC/MS test must confirm the presence of that specific metabolite.
In her appellate brief, Carmen argues that LabCorp didn't test for that specific metabolite and thus Labor Max failed to prove Gary's presumptive impairment. The final LabCorp test printout showed the presence of "Carboxy THC" at 189 nanograms per milliliter. Dr. Long referenced that result as showing 189 nanograms per milliliter of "THC-COOH." Neither the test printout nor Dr. Long's testimony specifically referenced delta-9-tetrahydrocannabinol-9-carboxylic acid by that name.
There's a reason Kansas law required testing for a specific marijuana metabolite: that's the metabolite that federally certified drug-testing labs are required to test for. And "THC-COOH" is just a commonly used reference for delta-9-tetrahydrocannabinol-9-carboxylic acid. See Evans, Drug Testing Law Tech. & Prac. § 4:144 (2018). So LabCorp, a federally certified lab, tested for the marijuana metabolite specified in the Kansas statute, and Dr. Long's testimony that the lab reported 189 nanograms per milliliter of *803"THC-COOH" confirms that. LabCorp's choice to list the result as "Carboxy THC" rather than writing out the full scientific name doesn't affect the result here.
Because the Board improperly excluded the LabCorp test result, we reverse the Board's decision and remand the case to the Board for further proceedings.