MILLER, J.

Employer Fleetguard, Inc. appeals and claimant Beverly Mannes cross-appeals from a district court ruling that affirmed in part and reversed in part a decision of the workers' compensation commissioner, and remanded the matter to the agency for consideration of whether Mannes suffered a cumulative injury to her neck and shoulders and when any such injury manifested. We affirm the district court.

I. Background Facts and Proceedings.

Mannes has been employed by Fleetguard since 1972. Mannes has worked as both a product assembler and a forklift operator. Between 1975 and the present, Mannes has suffered back injuries, a neck injury, a foot injury, and a rotator cuff tear. She has been diagnosed with carpal tunnel syndrome in both her left and right upper extremities, and with multiple sclerosis.

In October 2001 Mannes filed four petitions for workers' compensation benefits. The first three petitions collectively asserted Mannes had suffered an overuse injury to both arms on or about May 16, 2000. The fourth petition alleged Mannes had suffered an overuse injury to her neck and shoulders "[o]n or about" September 1, 2001.

Following an arbitration hearing, the deputy workers' compensation commissioner concluded Mannes had suffered a compensable injury to her arms on May 16, 2000. After reviewing the various medical opinions regarding the extent of Mannes's permanent impairment, the deputy accepted the opinion of the defendant's expert, Dr. Martin Rosenfeld, that Mannes's bilateral arm injury resulted in a two-percent whole body permanent impairment.

The deputy further determined Mannes had not sustained a work-related injury to her neck and shoulders on September 1, 2001. While apparently accepting that Mannes had suffered "cumulative trauma" to the area on September 1, 2001, the deputy determined "[t]here is nothing in the record that indicates that any event occurred on September 1, 2001 relative to this injury." The deputy rejected Mannes's contention that she should recover since September 1 was a "fair approximation" of her true injury date, which lay somewhere between July 24 and September 12, 2001. Relying on the prior agency decision of Kies v. Heritage Furniture, Inc., No. 1291174 (App. Dec. May 7, 2002), the deputy concluded Fleetguard lacked sufficient notice of an alternate injury date.

Mannes appealed, and the workers' compensation commissioner affirmed and adopted the deputy's decision. Mannes then filed an application for rehearing, asserting the Kies decision had been effectively overruled by the Iowa Supreme Court's opinion in University of Iowa Hospitals Clinics v. Waters, 674 N.W.2d 92 (Iowa 2004). In its ruling on Mannes's application, the agency found Waters to be distinguishable on its facts, and accordingly concluded itdid not overrule Kies. In rejecting Mannes's assertion that the deputy should have considered alternate injury dates, the agency noted,

[A]ny date alleged to be a date of injury should have some nexus with the governing rules and with what is being asserted in the claim. Defendants are not required to defend against claims that are not made. They are required to defend against the claim that was expressly made and reasonably anticipated variations of that claim, particularly where discovery or other information discloses that the date of the injury alleged in the petition is not necessarily the totality of the claim.

Mannes filed a petition for judicial review, asserting a number of errors by the agency. Relevant to this appeal, she contended the agency erred by failing (1) to decide whether she had a suffered a cumulative injury to her neck and shoulders and to consider alternate manifestation dates for the alleged injury, and (2) to consider evidence relevant to her impairment rating.

Following hearing, the district court rejected Mannes's contention that the agency failed to consider pertinent evidence relating to her impairment rating for her bilateral arm injury. However, the court agreed with Mannes's contention that the agency erred when it failed to determine whether she suffered a cumulative injury to her neck and shoulders, and when any such injury manifested, "rather than simply saying [she] picked an inappropriate date." The court also rejected Fleetguard's assertion that it would be prejudiced by an alternate injury date. The court accordingly reversed the agency decision that Mannes had not established a cumulative injury to her neck and shoulders, and remanded the issue to the agency for a determination regarding whether and when a cumulative injury occurred.

Fleetguard appeals. It contends the agency properly determined that Mannes did not establish an injury to her neck and shoulders on September 1, 2001, and that the court erred in reversing and remanding the matter to the agency because there was not adequate notice of any alternate injury date. On cross-appeal, Mannes contends the court erred in affirming the agency's two-percent permanent impairment rating for her bilateral arm injury. She contends the agency inappropriately ignored lay and expert testimony favorable to her position. She also appears to challenges the validity and reliability of Dr. Rosenfeld's impairment rating, and to assert that when all the relevant evidence is considered the agency's impairment rating is not supported by substantial evidence.

II. Scope and Standards of Review.

Iowa Code chapter 17A governs judicial review of decisions made by the workers' compensation commissioner. Iowa Code § 86.26 (2003). The district court acts in an appellate capacity to correct errors of law on the part of the agency. Grundmeyer v. Weyerhauser Co., 649 N.W.2d 744, 748 (Iowa 2002). A party challenging agency action bears the burden of demonstrating the action's invalidity and resulting prejudice. Iowa Code § 17A.19(8)(a). This can be shown in a number of ways, including proof the action was legally erroneous; unsupported by substantial evidence in the record when that record is viewed as a whole; or otherwise unreasonable, arbitrary, capricious, or an abuse of discretion. See id. § 17A.19(10).

On judicial review, we are bound by the agency's findings of operative facts, so long as those findings are supported by substantial evidence in the record when the record is viewed as a whole. See id.; Excel Corp. v. Smithart, 654 N.W.2d 891, 896 (Iowa 2002). In contrast, for those issues involving the agency's interpretation of the law, we determine whether the agency's interpretation was erroneous. See Clark v. Vicorp Rests., Inc., 696 N.W.2d 596, 604 (Iowa 2005) (citing Iowa Code § 17A.19(10)(c), (11)(b); Mycogen Seeds v. Sands, 686 N.W.2d 457, 465 (Iowa 2004)). In such instances, we may substitute our interpretation for the agency's. Id. However, when the challenge relates not to the agency's fact findings or interpretations of law, but to its ultimate conclusions, it is a challenge to the agency's application of law to the facts. Meyer v. IBP, Inc., ___ N.W.2d ___, ___ (Iowa 2006). In such cases "the question on review is whether the agency abused its discretion by, for example, employing wholly irrational reasoning or ignoring important and relevant evidence. Id. at ___ (citing Iowa Code § 17A.19(10)(i), (j)).

III. Cumulative Injury.

A cumulative injury is one that develops over a period of time. McKeever Custom Cabinets v. Smith, 379 N.W.2d 368, 373 (Iowa 1985). For the purpose of computing workers' compensation benefits, a cumulative injury occurs when "the `disability manifests itself.'" Oscar Mayer Foods Corp. v. Tasler, 483 N.W.2d 824, 829 (Iowa 1992) (citation omitted). "`Manifestation' is best characterized as `the date on which both the fact of the injury and the causal relationship of the injury to the claimant's employment would have become plainly apparent to a reasonable person.'" Id. (citation omitted). The agency is given "a substantial amount of latitude" in determining the date of manifestation, because it is an inherently fact-based determination. Id. Thus, when the agency determines a date of manifestation, that determination will be upheld so long as supported by substantial evidence. Id. at 830.

Here, however, the agency did not determine a date of manifestation, or even in fact whether a cumulative injury occurred. Rather, the agency concluded Mannes had not established that a cumulative injury manifested on September 1, 2001, and declined to consider whether Mannes had established a cumulative injury manifesting on a different date because it determined Fleetguard had not been sufficiently apprised of any alternate date. Determining whether an employer was sufficiently informed of the alleged basis for the employee's claim for compensation, via the employee's petition and the subsequent discovery process, is a matter within the agency's discretion. Waters, 674 N.W.2d at 96. Accordingly, our review of this issue is for an abuse of discretion. Id.

As recently noted by our supreme court, "workers' compensation law is for the benefit of working men and women, `and should be, within reason, liberally construed.'" Id. For that reason, petitions for workers' compensation benefits are not to be judged by technical rules. Id. at 96-97. Rather, the question is whether the employer "is . . . afforded a substantive right to be at least generally informed as to the basic material facts upon which the employee relies as a basis for compensation." Id. at 97 (citation omitted).

Stated another way,

due process requires that a party "be informed somehow of the issue involved in order to prevent surprise at the hearing and allow an opportunity to prepare. . . . . The test is fundamental fairness, not whether the notice meets technical rules of common law pleading."

Tasler, 483 N.W.2d at 828 (citation omitted). "A showing of prejudice is essential to establishing a due process violation." Id.

In concluding Mannes had not established that she had sustained a cumulative injury to her neck and shoulders on September 1, 2001, the agency relied solely on the fact Mannes had not pled an injury date other than September 1, 2001, and its determination the injury date was never "at issue." The agency's reasoning in this case is relatively consistent with the reasoning it employed in the Keis decision. However, to the extent Keis stands for the propositions relied upon by the agency, it is inconsistent with prevailing case law.

Given the liberal view of pleading in workers' compensation matters, it would be inappropriate to deny Mannes's claim of a cumulative injury simply because she pled in her petition that the date of manifestation was "[o]n or about" September 1, 2001, and did not plead multiple injury dates or otherwise allege specific alternative manifestation dates prior to the arbitration hearing. Concluding Fleetguard was deprived of due process on this limited basis, without determining whether Mannes suffered a cumulative injury, and if so when the injury manifested, was an abuse of discretion.

The agency should have engaged in "a separate analysis of each element" of Mannes's cumulative injury claim. See Meyer, ___ N.W.2d at ___. This includes an inquiry into whether Mannes established a cumulative injury, and the date when any such injury manifested. See id. at ___. Once these determinations were made, the agency could then look to the pleadings and the evidence disclosed to Fleetguard during the discovery process to determine whether Fleetguard was sufficiently apprised of the manifestation date of the cumulative injury. See Tasler, 483 N.W.2d at 828.

We recognize the agency's failure to engage in a thorough analysis of each element does not require reversal unless the agency's error prejudiced Mannes. See Iowa Code § 17A.19(8)(a). We conclude such prejudice was shown here, as a review of the record reveals the evidence available to Fleetguard prior to the arbitration hearing was sufficient to put it on notice of a manifestation date other than September 1, 2001. We first note the pleading itself alleged a manifestation date of "on or about" September 1. In addition, Fleetguard was aware Mannes was claiming a repetitive use injury to her neck and shoulders which became symptomatic in February 2001, when she began a new position at Fleetguard, and for which she consistently sought treatment, beginning July 24, 2001. The pre-hearing evidence also indicated Mannes received medical information tying her neck and shoulder condition to her work at Fleetguard sometime between October 11, 2001 and January 9, 2003.

Under this record it is possible that, if the agency determines Mannes established a cumulative injury, the manifestation date of the injury will be one of which Fleetguard had sufficient notice. Accordingly, the district court did not err in reversing the agency's determination that Mannes failed to establish a cumulative injury to her neck and shoulders, and remanding this matter to the agency for a determination of whether a cumulative injury occurred, and if so when the injury manifested.

However, it would be inappropriate to conclude, as did the district court, that Fleetguard would not have been prejudiced if the agency had found a manifestation date other than September 1. Resolving the question of whether any due process violation would occur is a fact-based determination dependent upon what date of manifestation, if any, is found by the agency. Accordingly, it is a matter to be determined by the agency upon remand if the agency determines that Mannes has established a cumulative injury, and if so, determines a manifestation date.

IV. Impairment Rating.

Mannes contends the district court erred in affirming the agency's two percent whole body permanent impairment rating because the agency either failed to consider or rejected without explanation lay and expert evidence demonstrating that Mannes had suffered "a far greater loss of function in her arms." She also appears to challenge the weight which may reasonably be attributed to Dr. Rosenfeld's opinion that Mannes suffered only a two percent whole body impairment, and the sufficiency of the evidence to support the agency's impairment rating. Before we explain our rejection of Mannes's claims, we find it necessary to give a summary of the relevant evidence.

On January 4, 2001, Dr. Clarence Carlson, Mannes's treating physician, sent a letter to Fleetguard which stated that he had been treating Mannes for bilateral wrist and hand pain, that Mannes was at maximum medical improvement, that he was imposing permanent restrictions of "[n]o forceful turning with either hand, avoid firm grasping with either hand," that "it would be best if [Mannes] was not subjected to highly repetitive tasks that would involve flexion or extension of her wrists," and that Mannes "can work in the warehouse area however . . . [she] should not be allowed to do water testing or line work because of her ongoing restrictions." On January 9, 2001 Dr. Carlson provided a written statement that "Mannes has no permanency disability rating assigned."

In January 2003 Mannes was evaluated by Dr. Elizabeth Stoebe. In a detailed report, Dr. Stoebe utilized the American Medical Association's The Guides to the Evaluation of Permanent Impairment (AMA Guides) to assign Mannes's bilateral arm injury a sixteen percent whole body permanent impairment rating. Dr. Stoebe also assigned a two percent "pain impairment" for all of Mannes's injuries, which included the alleged cumulative injury to her neck and shoulders and a foot injury that is not in issue in this matter.

At Fleetguard's request, Dr. Keith Riggins reviewed Dr. Stoebe's report. Dr. Riggins found the report "to be disjointed and very difficult to follow," and he was "frequently unable to identify findings upon which impairment ratings are based." Dr. Riggins opined that Dr. Stoebe did not appropriately use the AMA Guides. Based upon the range of motion restrictions in Dr. Stoebe's report, Dr. Riggins recalculated individual impairment ratings to the right and left wrists and thumbs and arrived at figures lower than those in Dr. Stoebe's report. Dr. Stoebe then recalculated her impairment ratings in light of Dr. Riggins's report. Although neither Dr. Riggins's review nor Dr. Stoebe's recalculation assigned a whole body impairment rating for the bilateral arm injury, Mannes asserts, and Fleetguard does not dispute, that under the AMA Guides conversion and combination of the individual impairment ratings assigned by the two physicians would result in a whole body impairment rating of somewhere between twelve and fourteen percent.

Mannes was also evaluated by Dr. Rosenfeld in January 2003. His report, issued after reviewing "a massive amount of information regarding" Mannes and conducting a physical examination, stated:

My impression is that she has overuse symptoms of her hands, resulting in limitation. . . .

In using the AMA Guides, I feel that because she does have limitations, the upper extremity hand problems are rateable and should be rated at 2% to each upper extremity because of the hand pain problems and necessity to avoid repetitive use.

. . . .

In using the combined values table and conversion tables, the 2% to each upper extremity would total 4% to the upper extremity, coverts to 2% body as a whole.

Finally, Mannes offered lay testimony about the adverse effects of her bilateral arm injury. The testimony, which included assertions the injury had resulted in pain, difficulty in holding objects, and loss of dexterity, extended to effects not necessarily within the contemplation of the AMA Guides.

The deputy found Dr. Stoebe's initial report to be "disjointed and difficult to follow." While recognizing that "[s]he improved a bit on this in her subsequent recalculation," he concluded that "since that recalculation was based upon the measurements that went into the first report that was so poorly drafted it too must be viewed critically, particularly in light of [Dr. Carlson's] rating of zero percent." The deputy acknowledged that, as Mannes pointed out, the AMA Guides have "limitations . . . in measuring disability," but determined that the greater weight of the evidence did not support her claim of a thirty-five percent whole body impairment. He then accepted Dr. Rosenfeld's two percent impairment.

As the foregoing illustrates, this case presents the classic situation of conflicting evidence. Dr. Stoebe's opinion and the lay testimony support a determination that Mannes suffered a whole body impairment rating greater than two percent. Dr. Carlson's statement and Dr. Rosenfeld's opinion support a determination that Mannes suffered a whole body impairment rating of two percent or less. Depending on how it is viewed, Dr. Riggins opinion lends weight to either position. Although he reached an impairment rating greater than two percent, he did not examine Mannes but relied wholly upon Dr. Stoebe's range of motion measurements. Thus, the import of Dr. Riggins's impairment ratings depends on the weight and credibility assigned to Dr. Stoebe's report, a report of which Dr. Riggins himself was highly critical.

It is the role of the agency to determine the credibility of witnesses and the weight to be given to any evidence, and it may accept or reject an expert opinion in whole or in part. Sherman v. Pella Corp., 576 N.W.2d 312, 321 (Iowa 1998); Dunlavey v. Economy Fire and Cas. Co., 526 N.W.2d 845, 853 (Iowa 1995). Thus, the agency was free to accept the opinion of Dr. Rosenfeld, as supported by Dr. Carlson's statement, over that of Dr. Stoebe. It is not the role of the district court upon judicial review, or this court on appeal, to reassess the weight and credibility of any of this evidence.

Mannes, however, contends this is not simply a matter of the agency finding certain evidence was entitled to greater weight or credibility. Rather, she asserts the agency "rejected or disregarded material evidence [in the form of Dr. Riggins's evaluation and the lay testimony] without any stated reasons." McSpadden v. Big Ben Coal Co., 288 N.W.2d 181, 186 (Iowa 1980) (citation omitted). If the agency had indeed done so, this would provide a basis for remand. See id. However, upon review of the deputy's decision, which was adopted by the agency, we conclude the agency gave consideration to this evidence and adequately stated the reasons for its weight determinations.

As previously noted, the weight to be given to Dr. Riggins's opinion, which is set forth in the agency's finding of facts, is inextricably tied to the weight and credibility given to Dr. Stoebe's report. The agency's rejection of Dr. Stoebe's impairment ratings, in which the agency was critical of not only her impairment calculations but the measurements upon which they were based, was an implicit rejection of Dr. Riggins's impairment ratings. Similarly, in agreeing with Mannes that the AMA Guides have limitations in measuring disability, the agency implicitly recognized her testimony that her bilateral arm injury had affected her in ways not easily measured by the AMA Guides. It then indicated her testimony was simply not enough to establish the extent of impairment she claimed. In light of the foregoing, we cannot conclude the agency rejected or disregarded material evidence without stated reasons.

Mannes also challenges the amount of weight which may reasonably be afforded to Dr. Rosenfeld's opinion and Dr. Carlson's statement. She points out that Dr. Rosenfeld did not specify he was using the most current version of the AMA Guides, and did not provide a detailed basis for his impairment rating. She also points out Dr. Carlson provided "no basis" for his statement that "Mannes has no permanency disability rating assigned," and there is no indication Dr. Carlson even applied the AMA Guides in reaching his conclusion. She appears to contend that, when all relevant evidence is considered, the agency's impairment rating is not supported by substantial evidence. However, we find these contentions to be without merit.

The alleged shortcomings of Dr. Rosenfeld's report and Dr. Carlson's statement are factors which affect the weight to be afforded the evidence. See Sherman, 576 N.W.2d at 321 (noting the weight to be given to an expert opinion by the agency does depend on the accuracy of the facts relied upon by the expert as well as other surrounding circumstances). However, that weight determination is a matter entrusted to the agency. Id.; Dunlavey, 526 N.W.2d at 853. Although we might wish for more detail in both Dr. Rosenfeld's opinion and Dr. Carlson's statement, or question how a rating of two percent physical impairment could be accurate given the substantial permanent physical restrictions assigned by the medical experts, Mannes has not demonstrated that such evidence must be rejected out of hand. Cf. Stephenson v. Furnas Elec. Co., 522 N.W.2d 828, 831 (Iowa 1994) (noting courts are not bound by factual findings "as to which there is no room for difference of opinion among reasonable minds"); Fazio v. Brotman, 371 N.W.2d 842, 844 (Iowa Ct.App. 1985) (noting that when "[t]he testimony of a witness [is] so impossible and absurd and self-contradictory . . . it should be deemed a nullity by the court").

The agency's impairment rating must be upheld if it is supported by substantial evidence, which is "the quantity and quality of evidence that would be deemed sufficient by a neutral, detached, and reasonable person, to establish the fact at issue when the consequences resulting from the establishment of that fact and understood to be serious and of great importance." Iowa Code § 17A.19(10)(f)(1). Although we must engage "in a fairly intensive review of the record to ensure that the fact finding is itself reasonable . . . evidence is not insubstantial merely because it would have supported contrary inferences." Wal-Mart Stores, Inc. v. Caselman, 657 N.W. 493, 499 (Iowa 2003) (citation omitted). Stated another way, the question in this appeal is not whether the evidence supports a finding different than that made by the agency, but whether the evidence supports the finding actually made. Excel Corp., 654 N.W.2d at 896. Upon review of the record, we conclude it contains evidence sufficient to support the agency's impairment rating.

V. Conclusion.

The district court did not err in upholding the agency's impairment rating. Nor did it err in reversing the agency's determination that Mannes failed to establish a cumulative injury to her neck and shoulders, and remanding this matter to the agency for a determination of whether a cumulative injury occurred, and if so when the injury manifested. Once these determinations are made, the agency should look to the pleadings and the evidence disclosed to Fleetguard during the discovery process to determine whether Fleetguard was sufficiently apprised of any manifestation date found by the agency.

AFFIRMED.