OPINION

STONE, Judge.

Relator is appealing the final disciplinary order of the Minnesota Board of Medical Examiners (Board) finding him subject to discipline for overprescribing benzodiaze-pines to a patient in 1981-82.

Relator argues that insufficient evidence existed on the record to find him in violation of Minn.Stat. § 147.021, subd. 1(g) and (k) and that the Board relied upon evidence not on the record and upon evidence to which relator had no opportunity to respond.

FACTS

Doctor Roger C. Kollmorgen is a medical doctor licensed to practice in Minnesota and three other states. He obtained his undergraduate degree in pre-med and psychology in 1959; a Master of Arts degree in vocational rehabilitation counseling in 1961; and a medical degree in 1964. He was licensed in 1967. In 1974, he was awarded a Ph.D. in Clinical Psychology. • He was Board certified in Psychiatry in 1978 and was pursuing a law degree at the time of this appeal.

The incidence of misprescription at issue here occurred while Dr. Kollmorgen was Medical Director and Staff Psychiatrist with the Five-County Mental Health Center (formerly the Five-County Human Development Program) in Braham, Minnesota. He first began treating the patient in question in September of 1980. At their initial session, the patient complained of depression, irritability, confusion, anxiety, insomnia, crying spells, and hallucinations.

Dr. Kollmorgen prescribed 3 different non-addicting tricyclic antidepressants *487 [Norpramin, Elavil (Endep) and Pamelor] during October and November of 1980. Each was discontinued by the patient within 3 to 5 days, apparently due to difficulty in tolerating the medication and lack of short-term effectiveness. The doctor then began prescribing benzodiazepines (Serax, Verstran, Ativan, Restoril, Prazepam and Dalmane) which are fast-acting and well-tolerated, but potentially addictive drugs. Dalmane and Restoril are prescribed to induce sleep and Ativan is used to allay anxiety.

There is divided authority on the maximum dosages, length of prescription, degree of physician supervision and advisability of prescribing benzodiazepines for those who have habitually abused alcohol or drugs in the past.

Dr. Kollmorgen knew of his patient’s in-patient treatment for alcoholism which had ended shortly before her first appointment with him. Indeed, it was the in-patient program that had referred her to him upon her release. The patient assured Dr. Kollmorgen that she was abstaining from alcohol, though she did, in fact, resume her use while she was on the benzodiazepines. At least two prescriptions were made, or renewed over the telephone while the patient was away at school. No allegations were made of any discernible harm to the patient, but a hearing was held to determine, in part, whether the dosages authorized, the length of treatment, lack of supervision, and the patient’s history of chemical abuse made the course of treatment subject to discipline.

The matter came on for hearing before an administrative law judge. The AU recommended the charges be dismissed because the Board’s Discipline Committee had not met its burden of proof.

Oral arguments were heard by the Board which found Dr. Kollmorgen subject to discipline under Minn.Stat. § 147.021, subd. 1(g) (conduct harmful to the public) and (k) (unprofessional conduct) (1984). He was then reprimanded and fined $1,000.

From that determination, the doctor appeals.

ISSUES

1. Whether the Board’s decision was based upon an independent examination of the evidence?

2. Whether Dr. Kollmorgen was given sufficient notice of, and opportunity to respond to, the charges against him?

3. Whether the Board’s decision was supported by substantial evidence in the record?

ANALYSIS

Agency decisions are to be accorded substantial deference, Manufactured Housing Institute v. Pettersen, 347 N.W.2d 238, 242 (Minn.1984). Minn.Stat. § 14.69 provides the standards for judicial review of an agency decision. A reviewing court may:

affirm the decision of the agency or remand the case for further proceedings; or it may reverse or modify the decision if the substantial rights of the petitioners may have been prejudiced because the administrative finding, inferences, conclusion, or decisions are:

(a) In violation of constitutional provisions; or

(b) In excess of the statutory authority or jurisdiction of the agency; or

(c) Made upon unlawful procedure; or

(d) Affected by other error of law; or

(e) Unsupported by substantial evidence in view of the entire record as submitted; or

(f) Arbitrary or capricious.

Minn.Stat. § 14.69 (1986).

Respondent alleges the Board based its decision to a substantial degree upon its members’ own experience and expertise rather than solely upon the record before it.

An agency that wishes to utilize the “general, technical or scientific facts within [its members’] specialized knowledge,” is required to take official notice of that information, notify the parties in writing, and provide them an opportunity to respond to *488 the information in question. Minn.Stat. § 14.60, subd. 4 (1986). But where the agency is examining evidence already contained in the record, it may use the ‘‘experience, technical competence, and specialized knowledge” of its members to evaluate the evidence in the record before it. Id.

Several allusions are made to the Board’s reliance upon its own knowledge and expertise in the memorandum accompanying its decision.

The first such reference concerns the advisability of prescribing these drugs to alcohol dependent individuals:

Benzodiazepines are viewed by many in the medical profession as tax deductible martinis. It is the view of most specialists in chemical dependency and Alcoholics Anonymous that benzodiaze-pines should rarely be prescribed to persons who have or who have had chemical dependency problems. That is also the view of each physician who serves on this Board.

It is apparent from this statement as well as from the summary of the testimony on subsequent pages that the Board fully considered all the evidence before it on this question. All the experts expressed the view that these drugs should rarely or never be prescribed to individuals with a history of alcohol dependency. The most liberal testimony before the Board held that ben-zodiazepines should be cautiously prescribed in limited amounts for limited periods of time to persons with such histories. Even Dr. Kollmorgen admitted to reservations about use of these drugs on this particular patient, but argued it was a judgment call supported by their beneficial effects on the patient. Nonetheless, the great weight of the evidence supports the Board’s conclusion. Any expression of the members’ view is merely cumulative and, if error at all, is harmless and non-prejudicial.

The second reference is to the question of overprescription:

[t]he Board through its physician members has experience, technical competence, and specialized knowledge on this issue which leads it to conclude that even if it were proper to prescribe benzodiaze-pines to Ms. A. as testified by Dr. Abuz-zahab, Respondent clearly over-prescribed them to her.

Once again the Board carefully summarizes the evidence on the record before reaching its conclusion. The Physicians Desk Reference (PDR) is generally recognized as authoritative among physicians. It contains recommendations by drug manufacturers, approved by the FDA, concerning drugs and their usage. A deviation from the manufacturer’s recommendation is considered prima facie evidence of negligence in Minnesota, if injury is proven. Mulder v. Parke Davis & Co., 288 Minn. 332, 339, 181 N.W.2d 882, 887 (Minn.1970). The PDR describes the “usual” adult daily dosage for benzodiazepines at 30 milligrams. Dr. Kollmorgen prescribed 30 mg dosages but one prescription permitted the patient to take one or two additional pills if necessary, allowing her, in effect, to double or triple the recommended dosage.

Two doctors testified that a 60 or 90 mg dosage amounted to an overprescription. One doctor testified that 15 mg is often sufficient and 90 mg was “exceedingly high.” Relator’s expert testified that the maximum recommended dosage was 30 mg. He was unconcerned about the dosages prescribed to the patient, but he had apparently not seen the evidence of the prescription authorizing repeat dosages.

Thus sufficient evidence existed for the Board to have concluded that Dr. Kollmor-gen had overprescribed benzodiazepines to his patient. Even though the Board uses language which may be ambiguous it is clear that the committee members considered all the testimony and evidence before it. Using members’ expertise to evaluate the evidence is permitted by the statute; creating a “secret record” is not. The Board does not appear to have created such a record.

This conclusion is further supported by the Board’s more carefully worded conclusion to this question:

Even under the ‘liberal’ standards espoused by Dr. Abuzzahab, it is clear to the Board as it reviews the testimony and in light of its own expertise and *489 knowledge that Dr. Kollmorgen over-prescribed benzodiazepines to [his patient].

Any use of member expertise was merely cumulative or an aid to evaluating the substantial evidence of overprescription on the record and thus did not constitute reversible error.

II.

Dr. Kollmorgen contends that he was denied due process of law because the subject of telephone prescriptions was not raised until the staff reply brief to the AU, denying the doctor the opportunity to answer the allegations.

Dr. Kollmorgen improperly characterizes the question here. Whether or not he prescribed medication by telephone is not independent grounds for discipline. Rather the question of misprescription or overprescription embraces many subquestions which must be evaluated before an answer to the primary question can be given. The nature of the drugs prescribed, the addictive tendencies of benzodiazepines, the patient’s alcoholism, the length and amount of drugs prescribed and the amount of supervision involved are all relevant to the question of overprescription. Proper supervision of drug use is especially important where habit-forming drugs are prescribed; even more important where the patient has a history of alcohol abuse and the drug administered has properties and effects so similar to alcohol.

The question of the impropriety of telephone prescriptions was first raised, ironically enough, in response to testimony by the doctor’s own witness. Dr. Abuzzahab was asked about the advisability of allowing automatic refills of benzodiazepines in light of medical standards and the requirements of the Controlled Substances Act. He replied:

With a minimum required standard of practice I would say patients taking ben-zodiazepines should be seen before the prescription is refilled, except that a patient can come in and [if the doctor sees] the patient is doing okay you might occasionally miss a monthly appointment.

The propriety of telephone prescriptions was, therefore, raised by the doctor’s own expert in the context of addressing the general question of overprescription.

Dr. Abuzzahab had also previously touched upon the question of supervision in an opinion letter written before the AU hearing. In it, he stated his opinion about the propriety of the treatment in question. He issued this letter under the misimpression that Dr. Kollmorgen was seeing the patient on a regular basis and was not permitting automatic refills:

the prescribed benzodiazepines were cautiously dispensed, in limited quantities, under close supervision, without any automatic refills. This insured that the patient would be seen by her physician before medications were given, (emphasis added)

As the doctor’s own expert regarded supervision as relevant and material to the general question of overprescription, the doctor himself should have been aware of the importance of the issue. Additionally, the doctor entered into evidence an opinion letter which on its face indicated the proper standard of care in this case — which was not followed by Dr. Kollmorgen. The Board later chose to challenge the foundation of that opinion; Dr. Kollmorgen cannot now argue he had no notice.

Even if the evidence of the telephone prescriptions is thrown out, enough evidence remained for the Board to have found Dr. Kollmorgen subject to discipline.

III.

In reviewing an agency decision, this court will examine the entire record to determine whether the decision is supported by substantial evidence. Reserve Mining Co. v. Herbst, 256 N.W.2d 808, 824 (Minn.1977).

The Board considered several factors before reaching its conclusion that Dr. Koll-morgen had misprescribed benzodiazepines.

1. Despite a misstatement to the contrary, it is obvious from the Board’s memorandum that it evaluated the prescriptions to the patient in light of her history of *490 chemical dependency. The vast weight of the evidence tended to support the conclusion that benzodiazepines should rarely or never be prescribed to patients with substance abuse problems. Nevertheless, the Board gave Dr. Kollmorgen the benefit of the doubt on this question, and utilized his medical expert’s comparatively liberal view that “benzodiazepines may be prescribed to persons with a history of chemical dependency cautiously and for a limited amount for a limited duration.”

2. The PDR states: “the prolonged administration of Restoril (temazepam) is generally not necessary or recommended.” One of the Board’s medical experts testified that benzodiazepines should not be prescribed for an extended period of time. Another doctor testified that sustained prescription was unethical and unprofessional. Even Dr. Kollmorgen’s' expert stated use should be for a limited duration for patients with a history of dependency. The doctor himself testified that long-term usage had not been adequately studied by the medical community. No evidence was offered to support such prescriptions over an extended period of time — here about 15 months. The weight of the evidence clearly supports the Board’s conclusion.

3. Additional controversy existed over proper interpretation of the December 28, 1981 prescription of forty 30 mg Restoril pills. The evidence produced at the hearing was a carbon copy of the actual prescription. Dr. Kollmorgen maintains that he wrote “MRXI” meaning “may repeat times one.” The Board contended that the prescription read “MRXII” or “may repeat times two.”

It appears from the photocopy that the Board’s interpretation was correct. Dr. Kollmorgen was free to produce the original form if he felt the photocopy did not reflect the original, but his failing to do so did not amount to a shifting of the burden of proof. In any event, the Board clearly stated that they believed the doctor’s prescription was excessive even if they accepted the doctor’s testimony and his interpretation of the photocopy.

Even if Dr. Kollmorgen’s testimony were correct, “MR XI” still means that the daily dosage prescribed was twice the recommended maximum in the Physician’s Desk Reference.

The Board conducted an independent appraisal of the document and concluded it read MRXII — that the second line was a roman numeral II and not merely a “carbon shadow.” Sufficient evidence existed to support the conclusions of the Board.

4. Dr. Kollmorgen also challenges the conclusion that two prescriptions were issued by telephone. He cannot recall whether or not he issued such prescriptions, but testified that he may have. While the direct evidence here is somewhat weak, circumstantial evidence established that the patient’s extended absence for schooling coincided with a period for which there are no clinical notes or other evidence of appointments, and during which the patient testified she filled prescriptions in the community in which she was attending school.' No directly contradictory testimony was offered, and sufficient evidence was offered which, if believed, would support the conclusion of the Board.

5. In addition to the issues addressed above, evidence was offered as to whether or not Dr. Kollmorgen had fallen below the minimum standard of care in his profession.

One Board expert concluded:

In my opinion in reviewing the prescription, copies of prescriptions and copies of prescribing forms indicating the amount and quantity and dates of scheduled sleeping medications or Benzodiaze-pines I would feel that this is inappropriate and below the minim[um] professional levels for any patient, unless there were a particularly severe medical illness, a far advanced cancer, etcetera, no matter whether the person were chemically dependent or not chemically dependent.

Particularly at the age of this patient, I think that ongoing month by month use of Benzodiazepines on a daily basis even though within the recommended therapeutic dose, still is below the minim[um] *491 level of ethical and professional conduct. There should be, at least at the very minim[um], attempts made to wean the person from this psychological if not physical addiction to sleeping pills and Benzodiazepines. If not that at least entries in the chart indicating why an ongoing sustained month after month use of Benzodiazepines is being used.

The other option would be to obtain a second opinion from another psychiatrist, whether indeed the continued use of Ben-zodiazephines is indicated.

His analysis simply summarizes and corroborates the standards of practice articulated by the P.D.R., the doctor’s own witness and the various opinions discussed more fully above.

DECISION

Based upon the substantial record before it, the Minnesota Board of Medical Examiners was justified in disciplining Dr. Koll-morgen for misprescription of benzodiaze-pines where such prescriptions exceeded the recommended adult dosages, were given over an extended period of time with inadequate supervision to a patient with a history of substance abuse. The weight of the evidence adequately supports the disciplinary action taken.

Affirmed.