OPINION

SIMS, J.

A writ of review was granted in these proceedings to consider petitioner's contentions that the unscheduled rating of permanent disability upon which the award was predicated was not determined in a manner authorized by section 4660 of the Labor Code, and that the respondent Workmen's Compensation Appeals Board violated the provisions of section 5908.5 of the Labor Code because it failed to state the evidence upon which it relied in denying petitioner's petition for reconsideration. For reasons set forth below it is determined that the board complied with the provisions of section 5908.5 by referring to the referee's report on the petition for reconsideration in its decision. Nevertheless, the award must be annulled and the matter remanded because the record does not contain evidence to sustain the conclusion in the report, which was adopted and incorporated in the decision of the board, that the rating specialist was using his own judgment in determining the unscheduled percentage of disability upon which the referee and the board relied, and because the petitioner was denied the right to cross-examine the rating specialist whose judgment the witness apparently followed.

The basic facts are not disputed, and are taken from the referee's report on the petition for reconsideration as follows: Applicant, born November 22, 1923, was employed as an office manager for the Department of Motor Vehicles at Oceanside. He filed an application alleging a gastrointestinal condition as a result of his employment from February 1, 1953 to May 27, 1969.

The applicant first noted symptoms in 1957 when he had a hemorrhage with a diagnosis of duodenal ulcer. He was given treatment and was off work for three weeks. He had surgery in 1960 when two-thirds of his stomach was removed. He was off for three months. In 1962 further surgery was performed when a nerve was cut. He has not had any hemorrhages since. In 1964 he was appointed assistant manager. He experienced some difficulty in the change of jobs. There was no drastic change in his symptoms until May 9, 1969. On that date a man came into the office and threatened to bomb him. This upset him physically. He had diarrhea for six days and his stomach was upset. He was off work five-six days and under the care of Dr. Thomas. Between May 9 and his three-week vacation in August, he lost several pounds in weight. He had an apprehensive feeling in meeting people and an increase in the dumping syndrome. He felt reasonably well while on vacation. On his return to work he had an apprehensive feeling and he would have a tight feeling in his abdomen and a gastric problem. An irate customer would upset his whole system. He did not have this apprehensiveness prior to May 9, 1969.

The findings and award found that the applicant did not sustain an injury as a result of his employment from February 1, 1953 through May 27, 1969, but rather that he sustained an injury on May 9, 1969.

With respect to the principal question raised in these proceedings, [1] the report continues as follows: "The defendants finally contend that the recommended rating does not follow the procedure set forth by the Labor Code and the rules of the Workmen's Compensation Appeals Board for rating disabilities. The defendants contend that the 20% standard used by the Permanent Disability Rating Specialist was not a scheduled rating and was not based upon the judgment of the Rating Specialist. At the hearing held April 20, 1970, the Permanent Disability Rating Specialist testified that he had been giving a 10% standard rating based on similar instructions to those in this case. However, about five or six years ago he found that his co-workers were using a 20% standard. Since that time he has been using 20% for this type of case. The defendants contend that the rating specialist was not using his judgment but was using a schedule not established by the Administrative Director. However, it would appear that the rating specialist was using his own judgment. About five or six years ago he had found that his judgment was low and therefore since that time he has considered a disability based upon the instructions in this case to warrant a 20% standard rating."

(1) Section 5908.5 provides in pertinent part: "Any decision of the appeals board granting or denying a petition for reconsideration or affirming, rescinding, altering, or amending the original findings, order, decision, or award following reconsideration shall be made by the appeals board and not by a referee and shall be in writing, signed by a majority of the appeals board members assigned thereto, and shall state the evidence relied upon and specify in detail the reasons for the decision."

In interpreting the foregoing provisions it is now established, "... that if the appeals board denies a petition for reconsideration its order may incorporate and include within it the report of the referee, provided that the referee's report states the evidence relied upon and specifies in detail the reasons for the decision." ( LeVesque v. Workmen's Comp. App. Bd. (1970) 1 Cal.3d 627, 635 [ 83 Cal. Rptr. 208 , 463 P.2d 432 ]. See also Gaiera v. Workmen's Comp. App. Bd. (1969) 271 Cal. App.2d 246, 248 [ 76 Cal. Rptr. 656 ]; and cf. Evans v. Workmen's Comp. App. Bd. (1968) 68 Cal.2d 753, 755 [ 68 Cal. Rptr. 825 , 441 P.2d 633 ].)

(2) The true issue in this case is not the board's failure to set forth the evidence upon which it relied, but whether that evidence, as related in the referee's report as adopted by the board, in fact exists.

Section 4660 sets forth the criteria for determining the percentage of permanent disability suffered, and provides for an official schedule which shall be prima facie evidence of that percentage for each injury covered by the schedule. [2]

The record shows that the referee requested a rating for a 45-year-old manager with a compensation rate of $52.50, who on May 9, 1969, suffered the following injury: "Stomach condition resulted in necessity to avoid emotional stress or strain." The rating specialist recommended a rating of 21 1/2 percent amounting to 86 weeks of disability payments at the rate of $52.50 in the total sum of $4,515. It is conceded that the rating of 21 1/2 percent represented a rating of 20 percent adjusted for age and occupation.

At petitioner's request a further hearing was held to cross-examine the rating specialist. The rating specialist categorically denied that the rating was a scheduled rating. Therefore neither the referee nor the board could rely on the provisions of subdivision (b) of section 4660 (see fn. 2 above) which make a rating for a scheduled injury prima facie evidence of the percentage of permanent disability. (See Young v. Industrial Acc. Com. (1940) 38 Cal. App.2d 250, 255 [ 100 P.2d 1062 ].)

The board relies upon the general principle most recently promulgated in U.S. Auto Stores v. Workmen's Comp. App. Bd. (1971) 4 Cal.3d 469 [ 93 Cal. Rptr. 575 , 482 P.2d 199 ], as follows: "The decision is supported by substantial evidence if the degree of disability found by the referee is within the range of evidence in the record. It is not necessary that there be evidence of the exact degree of disability. ( Liberty Mut. Ins. Co. v. Industrial Acc. Com. (1948) 33 Cal.2d 89, 93 ....)" (4 Cal.3d at pp. 474-475. In addition to the cited case, see Hines v. Industrial Acc. Com. (1932) 215 Cal. 177, 188 [ 8 P.2d 1021 ]; Ford Motor Co. v. Industrial Acc. Com. (1927) 202 Cal. 459, 464 [ 261 P. 466 ]; Michon v. Workmen's Comp. App. Bd. (1971) 15 Cal. App.3d 917, 924 [ 93 Cal. Rptr. 476 ]; Janet v. Industrial Acc. Com. (1965) 238 Cal. App.2d 491, 497 [ 47 Cal. Rptr. 829 ]; W.P. Fuller & Co. v. Industrial Acc. Com. (1962) 211 Cal. App.2d 9, 20 [ 27 Cal. Rptr. 401 ]; and Vincent v. Industrial Acc. Com. (1955) 136 Cal. App.2d 679, 682 [ 289 P.2d 246 ].) In this record, however, there is no range of evidence. The rating specialist testified that the rating furnished to and used by the referee, which was subsequently approved by the board, was not a judgment rating, that it was completely and exclusively based on the written request of the referee, and that it was not necessary for him to resort to or review any portion of the file other than the request.

It is recognized, as set forth in Fidelity & Cas. Co. v. Workmen's Comp. App. Bd. (1967) 252 Cal. App.2d 327 [ 60 Cal. Rptr. 442 ], "Because of the multiple factors which often result from injury, many cases require a `judgment' rating, i.e., some degree of individual evaluation in applying the schedule to disabilities not precisely described therein. The Permanent Disability Rating Bureau is comprised of specialists who evaluate and recommend ratings." ( 252 Cal. App.2d at p. 332 .) The distinction between scheduled and judgment ratings is set forth by Hanna as follows: "Experience has shown that only a small percentage of actual cases fit into the precise terms of the California rating schedule. In fact, it is not practicable by means of a schedule to cover each and every possible disability, since the variety and combinations of factors are infinite. Consequently, one cannot consult the schedule as if it were a timetable with the expectation of finding every conceivable disability listed. The disabilities enumerated in the schedule are those which experience has shown to occur more commonly. These are called schedule disabilities and the rating derived therefrom are termed schedule ratings. All others, referred to as nonschedule ratings or judgment ratings, must be determined by comparing the condition under consideration with the schedule disability most nearly resembling it, by analogy to a schedule disability, or by comparison with the entire scheme of schedule disabilities. All such nonschedule ratings call for an estimate which is based in part upon the judgment of the rating expert. Probably something less than 10% of all cases rated are pure schedule ratings." (1 Hanna, Cal. Law of Employee Injuries and Workmen's Compensation (2d ed. 1970) § 11.01[6], pp. 11-11/11-12. See also Davis v. De Laval Turbine and Fireman's Fund Insurance Company (W.C.A.B. 1970) 35 Cal.Comp.Cases 465, 466.)

In this case there is no evidence to show that the rating specialist made any comparisons, or drew any analogies with any schedule disability, or that he otherwise exercised his judgment. In fact, as noted below, his personal judgment had been that a lower rating was proper.

Nor can it be said that the referee and the board independently selected the rating from the facts in the case. In W.P. Fuller & Co. v. Industrial Acc. Com., supra , the court observed, "... if the commission may determine the question of permanency of injuries from evidence concerning the applicant's physical condition, without aid of expert testimony, there does not seem to be any reason why, where, as here, lay and medical evidence is given of applicant's physical condition arising from his industrial injuries and from systemic disease or condition, the commission from its vast experience may not determine the percentage of disability due to the industrial injuries and that due to the other causes, without the necessity of medical experts expressing percentages." ( 211 Cal. App.2d at p. 18 .) In this case there is no suggestion, nor has this court been referred to any evidence which would suggest that the referee or the board made an independent evaluation of the applicant's percentage of disability, as distinguished from reliance on the report of the rating specialist. (See Fidelity & Cas. Co. v. Workmen's Comp. App. Bd., supra, 252 Cal. App.2d 327, 333-334 .)

The rating specialist testified that he considered that "stress or strain" would go along together, one or the other, and that they were the equivalent of "stress" or "pressure." He acknowledged that originally he had been giving 10 percent for avoidance of stress and strain when it was the sole disability because in his judgment that was the appropriate rating for that type of disability.

He further testified, however, that at some undisclosed time when cross-examined by an applicant's attorney, who insisted that a 20 percent rating should be granted, he went to see his area supervisor, and was told to rate the limiting condition in question at 20 percent when it was the sole disabling factor. The witness explained that the 20 percent figure is what the rating specialists call a plateau, and all factors similarly given would get the same standard rating, a 20 percent; that the rating was established by the rating bureau five or six years prior to the hearing (April 1970) when the factor of avoiding emotional stress or strain became apparent in cases involving heart conditions; and that when avoidance of emotional stress or strain becomes an additional factor to whatever other rating is included in a heart case, or any other case in which that limitation of activity may occur, an additional 10 percent is added for that factor; that when taken by itself, as distinguished from when it is an additional factor to another disability, restriction occasioned by emotional stress has a standard or plateau of 20 percent; and that 20 percent would be the minimum rating for such limitation even if it were an additional factor to a disability rated at less than 10 percent.

The rating specialist stated that the instructions of the rating bureau with respect to the foregoing standards had been communicated to him verbally at meetings and he had no written record of them; that he did not know how his supervisor had arrived at the 20 percent figure; and that he merely followed his supervisor's statement because he was his superior in the rating bureau. There is, therefore, no evidence to support the finding in the referee's report that "it would appear that the rating specialist was using his own judgment."

At the conclusion of this testimony the petitioner requested that the hearing be continued to permit cross-examination of the area supervisor and the request was denied.

Reliance upon nonscheduled ratings on factors of work restriction, rather than physical impairment and limitation, in accordance with unpublished standards has been recognized as a common procedure in determining the percentage of disability. (See Davidson v. Workmen's Comp. Appeals Bd. (1966) 31 Cal.Comp.Cases 201, 202; and 1 Hanna, op. cit., § 11.02[3], pp. 11-19, fn. 24.) When such an unpublished plateau or guideline is determined "by comparing the condition under consideration with the schedule disability most nearly resembling it, by analogy to a schedule disability, or by comparison with the entire scheme of schedule disabilities" (see 1 Hanna, op. cit., § 11.01[6], as quoted above), it may be given full weight as a judgment rating. When, as here, however, there is no explanation of the manner in which the rating was evolved, other than upon the orders of a superior, it is analogous to an unpublished confidential report of the degree of disability.

In Young v. Industrial Acc. Com., supra , Mr. Justice Peters, then presiding justice of this court, observed, with the concurrence of one colleague, "At the oral argument it was disclosed that it is the practice of the respondent commission, after a case is closed, to submit the medical evidence to its own rating department, and to receive from that department a confidential report as to the degree of disability of the employee. In the present case, the award was based on the confidential report of May 11, 1939. Such practice cannot be approved. Neither the employee nor the carrier are given an opportunity to cross-examine the expert in order to demonstrate that he applied the wrong schedule to the particular case, or to show, as in the present case, that there is no schedule applying to the particular injury. The award must be based on evidence, and that means evidence legally and properly introduced. The commission, as trier of the fact, has no legal right to predicate its award upon the opinion of its own or any other expert, when such opinion does not appear in the record." ( 38 Cal. App.2d at p. 257 . See also Fidelity & Cas. Co. v. Workmen's Comp. App. Bd., supra, 252 Cal. App.2d 327, 332 .)

The situation is also analogous to Davis v. De Laval Turbine and Fireman's Fund Ins. Company, supra. There the rating specialist testified, "A 3 percent rating was given in this case because of a prior decision of the industrial accident commission to the effect that no rating less than 2 1/2 percent shall be given in a back case.... It has been the practice of the rating bureau ever since said decision to give a minimum rating of 3 percent where there is ratable back disability. In the absence of this established procedure, he probably would have rated the disability in this case as 1 percent flat." (35 Cal.Comp.Cases at p. 466.) The Appeals Board concluded, "It is apparent from this testimony that the 3% recommendation was based upon a legal conclusion and not upon the professional judgment of the rating specialist. His actual expert opinion is that the disability should be rated at 1%." ( Id. ) In short, there is no competent evidence to sustain the rating of 20 percent on which the final rating was predicated. (See Fidelity & Cas. Co. v. Workmen's Comp. App. Bd., supra, 252 Cal. App.2d 327, 335 ; Young v. Industrial Acc. Com., supra, 38 Cal. App.2d 250, 256 ; and Subsequent Injuries Fund of the State of California v. Industrial Accident Commission (1965) 30 Cal.Comp.Cases 336-343.)

Furthermore, it is clear that the petitioner was deprived of the effective right of cross-examination by the failure of the referee to grant it an opportunity to cross-examine the supervisor in the rating bureau who presumably did have knowledge of true criteria used to establish the unpublished plateau or standard. (See Allied Comp. Ins. Co. v. Ind. Acc. Com. (1961) 57 Cal.2d 115, 121 [ 17 Cal. Rptr. 817 , 367 P.2d 109 ]; Fidelity & Cas. Co. v. Workmen's Comp. App. Bd., supra, 252 Cal. App.2d 327 ; and Young v. Industrial Acc. Com., supra, 38 Cal. App.2d 250, 257 .) This is not to say that the supervisor must be called in every case where there is an unpublished plateau or standard. When the rating specialist can testify that the percentage was determined by use of the proper criteria, which have been referred to above, [3] the referee and the appeals board may apply the recommended rating as a judgment rating. Here there is no such evidence.

The award is annulled and the matter is remanded to the appeals board for further proceedings.

Molinari, P.J., and Elkington, J., concurred.

[1] Petitioner's further contentions that the referee was without power to amend the application and find an injury on a specific date when the petition alleged continuous trauma over a period of time, and that an apportionment should have been made, have not been pursued before this court.

[2] Labor Code section 4660 provides in pertinent part: "(a) In determining the percentages of permanent disability, account shall be taken of the nature of the physical injury or disfigurement, the occupation of the injured employee, and his age at the time of such injury, consideration being given to the diminished ability of such injured employee to compete in an open labor market.

"(b) The administrative director may prepare, adopt, and from time to time amend, a schedule for the determination of the percentage of permanent disabilities in accordance with this section. Such schedule shall be available for public inspection, and without formal introduction in evidence shall be prima facie evidence of the percentage of permanent disability to be attributed to each injury covered by the schedule."

[3] The distinction may be exemplified by comparison with two cases which each party has requested the court to consider with the principal case. In State Compensation Insurance Fund v. Workmen's Compensation Appeals Board [ Thompson ] 1 Civil No. 29810, the instruction to the rating specialist read, "Peptic ulcer condition requires avoidance of emotionally stressful conditions as much as possible." The referee adopted the permanent disability rating recommended by the rating specialist which was based on a nonscheduled, oral, unpublished guideline of 20 percent for a person precluded from activity involving emotional stress or strain. As distinguished from the principal case [ Payne ], the rating specialist testified that he exercised his independent judgment by using the unpublished plateaus or guidelines and by comparing scheduled ratings for other disabilities with the stated disability. He fully explained the manner in which the guidelines were established and the reasons for his judgment.

In State Compensation Insurance Fund v. Workmen's Compensation Appeals Board [ Nelson ] 1 Civil No. 29811, the instruction was, "Heart condition limiting employee to semi-sedentary activities not involving emotional stress." The referee adopted the permanent disability rating recommended by the rating specialist which was predicated upon a scheduled 60 percent rating for a "semi-sedentary" limitation, plus 10 percent for preclusion from work involving emotional stress, or a total unscheduled rating of 70 percent. Here again the rating specialist (the same as in Thompson, supra ) testified that the rating was made in the exercise of his independent judgment. He explained the comparisons he had made with the rating schedule. He acknowledged that the question of the rating for preclusion from activities involving emotional stress had been discussed among personnel of the rating bureau, but it was not established as in the principal case [ Payne ], and as in Thompson that there was an unpublished plateau or guideline.

An alleged discrepancy in the testimony of the witness in Thompson and Nelson was not brought out by cross-examination in the later proceeding, does not in fact exist when the two scheduled ratings to which reference was made are compared, and, in any event, would have been a discrepancy for resolution by the trier of fact.

In both Thompson and Nelson the referee was warranted in finding that the rating specialist exercised his independent judgment. The reasons for that judgment were fully explored, and could but go to the weight to be given his testimony a question to be resolved by the referee and the board. In each case a petition for review has been denied this day.