Terry M. Kinney, Office of the United States Atty., Crim. Div., Chicago, IL, Barry Rand Elden, Chief of Appeals, Office of the United States Atty., Crim. Appellate Div., Chicago, IL, Joseph D. Wilson (argued), Dept. of Justice, Crim. Div., Washington, DC, for U.S.

John T. Theis (argued), Chicago, IL, for George Neville.

James D. Montgomery (argued), Thomas C. Marszewski, Montgomery & Associates, Chicago, IL, for Juliet S. Dumlao.

Gary J. Ravitz (argued), Chicago, IL, for Betty Sanchez.

Before CUMMINGS, WOOD, Jr., and EASTERBROOK, Circuit Judges.

HARLINGTON WOOD, Jr., Circuit Judge.

These consolidated cases provide an illustration of the elaborate schemes and risks some individuals--including a physician with hard-earned professional credentials--are willing to undertake to attain the dubious benefits of the drug trade. Defendants-appellants Betty Sanchez, George Neville, and Juliet S. Dumlao were convicted of conspiring over a four year period to possess and distribute over 60,000 Dilaudid tablets in violation of 21 U.S.C. § 846.1 Neville and Sanchez were each also found guilty on several counts of distributing Dilaudid and acquiring Dilaudid by fraud in violation of 21 U.S.C. §§ 841(a)(1) and 843(a)(3), and of one count of using a minor to violate the same statute. Dumlao was also convicted on thirteen counts of acquiring Dilaudid by fraud. Sanchez and Neville each were sentenced to life imprisonment, while Dumlao received 235 months, to be followed by three years of supervised release.

On appeal various issues are raised by one or more of the appellants. These arguments include sufficiency of the evidence, the omission and inclusion of certain instructions, the trial court's refusal to allow a public authority defense, whether the Speedy Trial Act was violated, and whether the amount of drugs was calculated correctly for sentencing purposes. After presenting the background of the case, we address each of these arguments in turn.

The offenses at issue stem from a scheme to obtain a substance known to the pharmaceutical trade as Dilaudid. To a chemist, Dilaudid is a tablet containing hydromorphone, a highly addictive narcotic. To a physician, Dilaudid is a useful drug in legitimate efforts to manage a patient's severe pain. On the streets, however, where it is frequently abused by illegitimate users, Dilaudid is known by a simpler name: "drugstore heroin."

As a synthetic narcotic of consistent quality, Dilaudid is much sought after as a street drug; users typically crush the pill into water, heat the mixture, and inject it. Dilaudid tablets are available in one, two, three, and four milligram dosages, but the more potent four milligram tablet is the most desired on the underground market. The legitimate price for Dilaudid at a pharmacy is between 29 cents and $1 per pill, but the same single pill on the streets of Chicago fetches between $25 and $45, and even more elsewhere.

Betty Sanchez and George Neville were familiar with these realities and became experts in Chicago's underground Dilaudid business. At least since the early 1980s, each developed and maintained extensive arrangements to obtain and resell Dilaudid, at least at the wholesale level. These arrangements were detailed, for obtaining Dilaudid once, let alone securing a steady supply, is no easy task. Dilaudid is available only by prescription, and even then is generally reserved for patients who are terminally ill or who have been hospitalized with severe trauma. Furthermore, in Illinois, a physician who wishes to prescribe a Schedule II narcotic2 like Dilaudid must use a triplicate prescription form provided by the Illinois Department of Alcohol and Substance Abuse (IDASA). Once this form is filled out, the prescribing physician keeps one copy, the pharmacy filling the prescription retains another, and the third is forwarded to IDASA for its work in tracking the use and illegal diversion of narcotics. Prescriptions written on such forms must be filled within two days of their date of issue.

To evade these restrictions and safeguards, Sanchez and Neville each identified and trained a group of "professional patients." These individuals, often indigent and/or illiterate, usually had serious past or present medical conditions. As a result, with some coaching in how to complain of severe pain and how to prevent a doctor from prescribing a lesser medication (by claiming an allergy to aspirin or codeine, for example) these "patients" were often able to obtain prescriptions for Dilaudid.

For Neville, obtaining the drug through professional patients was a fairly straightforward process. He typically had four or five professional patients make appointments, have prescriptions filled, and bring him the Dilaudid, all for a flat fee. Sanchez's efforts were more involved. Instead of relying on the patients, Sanchez made the appointments for her group, often obtaining a doctor's name from the yellow pages or by purchasing names of certain doctors from Neville or others. She then enlisted various friends and relatives to transport the patients to the doctor's office and to the pharmacy. Some of the patients were unable to talk, so Sanchez, posing as a concerned friend or relative, wrote letters for them to give to their doctor. These letters described some terrible chain of events often involving foreign travel and lost medications, and invariably suggested that the solution to the problem was a Dilaudid prescription. Otherwise Sanchez accompanied the patient and told the story herself. Still at other times she simply found a doctor, displayed her own scars, and complained of pain to obtain a prescription on her own.

These various ruses were somewhat successful; between the summer of 1987 and early 1988, for example, Sanchez obtained sixty-four prescriptions for Dilaudid. While false IDs were used to circumvent the triplicate prescription form system, gaining the help of the other necessary party--the doctor--was apparently more difficult. Sanchez's frequent search for new doctors indicates that many were suspicious; others apparently either got smart, got scared, or just got out. Meanwhile Sanchez, Neville, and others discussed and "traded" names of doctors they found helpful. This was done until they concluded that one way or another the doctor was "burned out"; they then moved on to another.

The government alleged that sometime in 1988 the Sanchez and Neville group discovered Dr. Juliet Dumlao, who then joined their conspiracy. Dumlao, a native of the Philippines, was engaged in the private practice of internal medicine in Oak Park, Illinois. She was fifty-two years old at the time of trial, and had practiced for several years. Her practice was rather notable in that she emphasized the religious and spiritual aspects of health; she described her medical philosophy to the jury somewhat ambiguously as the "pioneer holistic approach." Unfortunately, her practice was also notable for its lack of financial success. Dumlao had little or no net income and a negative financial worth. Her tax returns for the years 1986-1992 were admitted into evidence; only once in that seven-year period did she report net income more than $3,500, and three other years show a net loss.

Nearly all of Dumlao's patients during the period of the alleged conspiracy were individuals connected with the Sanchez and Neville group. Sanchez acted as a gatekeeper, controlling access for approximately thirty-five professional patients. During the alleged four-year conspiracy, these thirty-five patients received more than 600 prescriptions for over 63,000 tablets of Dilaudid. Every one of these scripts called for the four milligram dosage, and more than sixty of them were left undated, thus avoiding the two day limitation imposed by IDASA.

The jury also heard several facts and opinions about Dumlao's prescribing practices. For example, though all the patients were seen for an initial visit, and at least appeared to have serious medical conditions, many did not appear again but got more Dilaudid prescriptions anyway. Medical experts testified that Dumlao's documentation of the initial patient visits was minimal. Rarely, if ever, did Dumlao conduct tests to verify their claims of serious illness. Testimony also indicated that the amounts and dosages of Dilaudid prescribed did not vary with the condition, need, or weight of the patient. Sometimes a Dilaudid prescription was even contraindicated (if a patient complained of respiratory problems, for example, which some did). Patients were also sometimes allowed or instructed to share their medication, even though the medical profession views this to be an unacceptable practice. About a third of Dumlao's patients claimed to be allergic to codeine (an ingredient in less potent pain medications), while the incidence of such an allergy in the general population is about one in 3,000.

Occasionally Dumlao would secure a signed note from a patient--addressed "to whom it may concern"--in which the patient would promise not to sell their medication, or would promise to enroll in a methadone program, but these were never followed up in any way. Meanwhile investigators from IDASA left messages for Dumlao stating that they wished to speak to her about professional patients, but Dumlao never returned their calls.

Dumlao's answer to all this was that she was a trusting doctor who was fooled by the schemes of Sanchez. Specifically, she believed Sanchez was a government agent who was in charge of several persons involved in a federal government witness protection program.

Sanchez was indeed quick to claim federal authority. She told various people including Dumlao that she was a "federal agent"; sometimes she claimed to work for the FBI, sometimes the DEA. Sometimes she was purportedly trying to expose bad doctors, sometimes bad pharmacists. The only consistency in all these claims was their untruth. Sanchez at one point had indeed been a paid informer for the DEA, but had been deactivated and was considered only an unpaid "source of information" after 1986.

Relying on her claimed authority, Sanchez wrote several notes and letters to Dumlao explaining that the government, through the witness protection program, had charged her with the care of several patients who were to testify against George Neville (actually her coconspirator). She told Dumlao that for various reasons related to the witness protection program these patients needed Dilaudid or were unable to make appointments. Many were supposedly waiting for departure to safe houses in Australia, South Africa, and other distant locales; others had just returned to the United States but needed new prescriptions because their Dilaudid had been confiscated by customs officials. Once Sanchez even claimed this happened to an American returning on a flight from Hawaii. Dumlao's receptionist testified that she warned Dumlao that Sanchez was not to be trusted, but Dumlao simply smiled in response.

The supposedly official notes and letters Dumlao received from Sanchez were handwritten, on steno pad or other nonstationery paper, and contained several egregious errors. One typical letter read as follows, with errors mostly uncorrected:

Write these prespition for the patience I am in Springfield waiting for the Governor to sign the check. I have to stay here because they gave me appointment for Friday at 1 oclock so I will stay here until then Write for me my medication all of it write the Dilaudid for 100

Write for Jerry Shaw and Carol Turner write it for Dialudid 150 ...

Write for Miguel Benevite he has to have his name on it For 100 Dilaudid ...

Do not keep Jim wait and write these all today

So he can come down here and bring my medication do not play no games

I have been down here since Tuesday I did know I had to have appointment I have your 1,200 with me. These people have to go because they testify and they have to be in the safe house

Two points may clarify this letter somewhat. First, the "Jim" in the letter refers to Sanchez's supposed son, Jim Morales, a fictitious name for one of Sanchez's drivers, who would take the prescriptions from Dumlao and then drive patients--or even Neville's minor son--to pharmacies to have them filled. Second, at one point Dumlao reported to IDASA that her triplicate prescription forms had been lost or stolen. These forms were later found, but her temporary inability to write prescriptions for restricted substances was apparently the "game" Sanchez suspected her of playing.

The jury saw several of these notes and letters. They offer a variety of scenarios and excuses. One that deals with many facets of the interaction between the group and Dumlao is reproduced below. Other than being typewritten, more comprehensive, and signed by a supposed associate of Sanchez's ("Agent Eugene Corbitt"), it is not different in style or substance than the others Dumlao received. Uncorrected3 except for margins and converted to single spacing, it reads:

To whom it may concern;

All these people are Federal witness in George Nevelle worker's cases and the people who work for him selling cocaine and for murder. They are giving Testimony which will convict and send them all to prison. We need all the evidence Mrs. Moralas oversee all of them and give them their Medication. We having been waiting on their medication so we can return them all to the safe house. Where they can get their mind all settled and not under stress, so they will be able to testify and go through the whold trail without maling mistakes. The trail has been held up for 72 days now. The Judge has held Mrs. Moralas for contempt of court over telling them lies. She says every day that they will get their medicalion which they have not got. You said when you got your Triplicate you would give every one of them their medication. You have caused this trail to be interupted and it all ready cost us 2 million in expenses....

We must return these people this week to the safe house. For their health and safety. They are not criminals themselfs. all these people need is their medication Rita Smith, Barbara McNamara, Dollie Rogers ... [the list continues with 26 other names].

....

We thank you for your cooperation with us. Mrs. Moralas and all the are not allowed to discuss anything about this trial with any one, and all the other witnesses are sitting targets for George and the Mafia to give order to have them killed. Especially Mrs. Moralas, Jeanette Nevelle, Norma Shrum, Goldie Gorham, Dollie Rogers, Rita Smith, Carol Turner, Edmond Donovan and Joyce

s/ Agent Eugene Corbitt.

Sanchez's letters often referred to payment, which though ostensibly coming from the government took place by Sanchez giving Dumlao envelopes containing cash. Jeanette Neville, the wife of George Neville and a patient of Dumlao's, testified that she had heard loud arguments between Sanchez and Dumlao concerning payment, and on one occasion heard Sanchez threaten Dumlao that if the prescriptions were not written, Dumlao would "go down." Faced with this warning, Dumlao supposedly responded that she could always return to practice in the Philippines. Other witnesses also testified they heard other arguments over payment arrangements.

After a six-week trial, the jury found for the government on nearly every charge in the indictment. The defendants appeal.

In challenging the sufficiency of the evidence, Dumlao bears a heavy burden. United States v. Gutierrez, 978 F.2d 1463, 1468 (7th Cir.1992). Our review in such cases is limited; it is not our function to reweigh the evidence, nor substitute our judgment for the decision of the jury. United States v. Hatchett, 31 F.3d 1411, 1416 (7th Cir.1994). We consider the evidence in the light most favorable to the prosecution, making all reasonable inferences in its favor, and will affirm the conviction so long as any rational trier of fact could have found the essential elements of the crime beyond a reasonable doubt. Jackson v. Virginia, 443 U.S. 307, 319, 99 S. Ct. 2781, 2789, 61 L. Ed. 2d 560 (1979). Reversal is warranted "only when the record is devoid of any evidence, regardless of how it is weighed, from which a jury could find guilt beyond a reasonable doubt." United States v. Garcia, 35 F.3d 1125, 1128 (7th Cir.1994) (quoting Gutierrez, 978 F.2d at 1468-69).

a. The conspiracy conviction

We first consider Dumlao's claim that the evidence was insufficient to support her conspiracy conviction. To sustain this conviction, the government must provide substantial evidence that (1) a conspiracy existed and (2) that Dumlao knowingly agreed to join it. United States v. Cabello, 16 F.3d 179, 181 (7th Cir.1994). We stated in United States v. Townsend, 924 F.2d 1385 (7th Cir.1991), that these elements may be proved entirely by circumstantial evidence. Id. at 1390 (citations omitted). The critical question, then, is whether the jury may reasonably infer that the necessary elements existed. Id.

Given these standards, there is ample evidence to support such an inference in this case. We have held that transactions in large quantities of drugs, prolonged cooperation between the parties, and "[e]vidence that the parties have standardized their transactions with one another" are all relevant factors in determining whether a conspiracy exists and whether one has knowingly joined it. See United States v. Kozinski, 16 F.3d 795, 808 (7th Cir.1994). These factors are present here. The jury heard evidence of Dumlao's many suspicious prescribing practices, including supplying quantities of Dilaudid far exceeding the needs of her purported patients. These practices continued over a period of four years. The jury also heard evidence of Dumlao's expensive, standardized charges-allegedly over $300 per prescription--which were always paid in cash, except for one bounced check.

Dumlao objects that these indicators are meant to apply only to what would normally be arms-length transactions between independent actors, not ongoing cooperative relationships such as those between a doctor and her patient. We have recognized the underlying rationale for this argument, which is that conspiracies, like business combinations, are often formed to reduce transaction costs. Townsend, 924 F.2d at 1394. It is true that these factors, standing alone, describe professional consultative relationships just as they indicate the possible existence of a conspiracy between people who are otherwise competitors. But this argument ignores the fact that these are relevant factors, not exclusive ones. This is another way of saying that their probative value often depends on context.

The jury had many contextual factors to consider in this case, especially as related to whether the interactions could be characterized as part of a doctor-patient relationship. The jury heard of futile attempts by the government to warn Dumlao about specific professional patients; such a warning is enough for most physicians to drop such patients immediately. Moreover, Dumlao claimed that she did not know that the drugs she prescribed were being used for illicit purposes, but yet she required some patients to sign "to whom it may concern" notes promising that they would not sell their medication. She knew that many patients had severe drug addictions, and also frequently left the date off her prescriptions so they could be filled at any time, thus evading regulatory laws. Aside from these facts, there is even a question as to what extent a doctor-patient relationship existed at all, since Dumlao met many patients only once, and dealt with Sanchez as a go-between thereafter. The jury could reasonably infer that Dumlao was in fact aware of the patients' intentions and the results of her own actions, and thus that an ordinary doctor-patient relationship did not exist.

Much evidence against Dumlao was independent of any potential argument about factors drawn from economic theory or the attributes of professional relationships. The jury also heard a recounting of shrill arguments between Sanchez and Dumlao over prescriptions and the payments for them. At times objects were apparently thrown or desks were cleared with a dramatic swipe of an arm. And while Dumlao asserted at trial that she ran a "fee for service" practice based on consultation time, even a patient/witness, Jeannette Neville, the wife of co-conspirator George Neville, testified that she understood that Dumlao's fees were charged per prescription, not per visit. These things are not features of usual medical practice or indicative of interactions with responsible government officials. As a further anomaly, Sanchez's letters virtually demanded prescriptions and instructed Dumlao in various ways, including the admonition "do not play no games."

Finally, in addition to all these factors, Dumlao's acceptance of Sanchez's many "government agent" notes and letters easily permits an inference of Dumlao's awareness of the scheme. The many typographical errors aside, merely the content of Sanchez's letters to Dumlao was a glaring indication that something--really, everything--was amiss. The government argued that in view of her long years of education and experience, as well as her ethical and legal responsibilities related to the distribution of narcotics, a physician in Dumlao's position would not be so easily deceived. Moreover, the jury also heard that Dumlao had little to say about these letters in her defense. Referring specifically to the error-filled letters on direct examination, her attorney asked her: "Did that signal anything to you? Did that make you suspicious at all?" This was Dumlao's reply:

Well, knowing the person that has written the letter and how thorough she was, and this as far as she was--I am convinced as far as she but a list a lot as far as their personality, type A, and having believed her story because some of the persons that she mentioned were family members of previous patients I had. It was easy for me to connect what they were working on.

And when she writes the letter--and I am thinking more of the patient that is involved for one thing. At the same time I did notice a few misspellings, but that didn't basically--have worked in America I have seen some grammatical errors for English-speaking people. It's common. Because myself I come from a foreign country. Grammatically we are--because it's not our language, they are emphasizing quite a lot.

So even a medical person may misspell vomiting with two Ts instead of one. It's not surprising for me, for a patient for me to misspell certain words, and that was the explanation for that.

The finder of fact believed another explanation, and we will not question what was in effect their determination of credibility. Viewing all these factors in the light most favorable to the government, we find that a rational jury could reasonably infer, beyond a reasonable doubt, that a conspiracy existed and that Dumlao knowingly joined it.

b. The fraud convictions

Dumlao also appeals her conviction on thirteen counts of knowingly or intentionally acquiring or obtaining a controlled substance by misrepresentation, fraud, deception, or subterfuge in violation of the Controlled Substances Act, 42 U.S.C. § 843(a)(3). First, like her defense to the conspiracy charge, Dumlao argues that she was not aware that she was unlawfully issuing prescriptions, and thus did not have the requisite intent and knowledge to perpetrate a fraud upon pharmacies. In sum, Dumlao claims she was not the source of the fraud, but its victim--"merely the conduit" for the misrepresentations. This is a variation of the argument addressed above in the conspiracy context, and goes no further here. The jury found otherwise, and had a rational basis for doing so.

Dumlao next argues that the evidence offered to prove some of these counts was insufficient. Seven of the thirteen counts were based on prescriptions for six patients, none of whom testified at trial. Citing the Fourth Circuit's decision in United States v. Tran Trong Cuong, 18 F.3d 1132 (4th Cir.1994), Dumlao claims that the lack of testimony from these patients is impermissible because it invites the jury to find guilt by association or pattern, rather than consider each count independently.

In Tran Trong Cuong, the government attempted to prosecute a physician for over 130 counts of distribution of a controlled substance. The evidence for eighty of these counts, however, consisted merely of copies of the prescriptions and an expert's testimony and exhibit summarizing a fraction of the patients' files. 18 F.3d at 1141. None of the twenty patients who received the prescriptions testified. Id. Finding this to be "a classic example of 'overkill' by the prosecution," the Fourth Circuit reversed Tran's convictions on the eighty counts for insufficiency of the evidence. Id. at 1142.

In a more recent case, however, the Fourth Circuit emphasized that the presence or absence of witness testimony is not the only criterion of sufficiency. In United States v. Singh, 54 F.3d 1182 (4th Cir.1995), the court stated that "[i]n Tran Trong Cuong, we reversed the convictions, not because the victims did not testify, but rather because their lack of testimony was not replaced by any substantive evidence." Singh, 54 F.3d at 1188. Indeed, while witness testimony is important, it is difficult to see why it should be the touchstone in such determinations, especially given the fact that many patients in this case had terminal illnesses or suffered from a variety of other misfortunes which might have precluded availability.

In any event, unlike Tran Trong Cuong, there was plenty of substantive evidence regarding these counts without witness testimony. Much of this information was individual and particularized. In addition to the copies of the specific prescriptions, the government produced Dumlao's medical files on the patients in question. Furthermore, the government's expert reviewed all of these files and concluded that none contained adequate justification for a Dilaudid prescription--indeed, he noted some files contained information making such a prescription improper. Moreover, as a general factor, the jury was aware that all of these nontestifying patients were associated with Sanchez and her scheme to obtain Dilaudid. With these additional items of information, the jury was not asked to infer guilt on these counts merely from the other improper prescriptions. Accordingly, we find the evidence sufficient to support the convictions rendered.

Dumlao next argues that the trial court erred in giving an "ostrich" instruction. Such an instruction informs the jury that actual knowledge and deliberate avoidance of knowledge are the same thing. United States v. Ramsey, 785 F.2d 184, 189 (7th Cir.) (suggesting the form of the instruction currently in use), cert. denied, 476 U.S. 1186, 106 S. Ct. 2924, 91 L. Ed. 2d 552 (1986). We review determinations concerning jury instructions for abuse of discretion, considering the evidence and all reasonable inferences therefrom in the light most favorable to the government. See United States v. Talkington, 875 F.2d 591, 596 (7th Cir.1989).

Despite an earlier determination that an ostrich instruction was not appropriate in this case, Judge Nordberg considered later testimony as well as recent cases from this court, particularly United States v. Smith, 995 F.2d 662, 674 (7th Cir.1993), cert. denied, --- U.S. ----, 114 S. Ct. 718, 126 L. Ed. 2d 683 (1994), and United States v. Stone, 987 F.2d 469, 470-71 (7th Cir.1993), and after hearing arguments on the issue at the charging conference, concluded that one should be tendered. The court instructed the jury on this issue as follows:

When the word "knowingly" is used in these instructions, it means that a defendant realized what he or she was doing and was aware of the nature of his or her conduct and did not act through ignorance, mistake or accident. Knowledge may be proved by the defendant's conduct[.] You may infer knowledge from a combination of suspicions and indifference to the truth. If you find that a person had a strong suspicion that someone had withheld some important facts, yet shut his or her eyes for fear of what he or she would learn, you may conclude that he or she acted knowingly.

Dumlao does not object to the content of this instruction, which has been repeatedly endorsed by this court. See, e.g., United States v. Jackson, 33 F.3d 866, 874 (7th Cir.1994), cert. denied, --- U.S. ----, 115 S. Ct. 1316, 131 L. Ed. 2d 197 (1995); Ramsey, 785 F.2d at 188-91. Rather, Dumlao claims that this instruction allowed the jury to find knowledge on the part of Dumlao if it inferred that Dumlao knew or should have known that the misrepresentations made to her were false; it allegedly criminalizes her mere negligence.

We have repeatedly held that "an ostrich instruction is appropriate when 'the defendant claims a lack of guilty knowledge and there are facts and evidence that support an inference of deliberate ignorance.' " United States v. Bigelow, 914 F.2d 966, 970 (7th Cir.1990) (quoting Talkington, 875 F.2d at 596), cert. denied, 498 U.S. 1121, 111 S. Ct. 1077, 112 L. Ed. 2d 1182 (1991). Whether the defendant's ignorance is deliberate is the key factor separating carelessness from avoidance. This is often a close question, but still a legitimate inference of willfulness is sufficient. Moreover, this inference need not be shown by overt or physical acts. "Evidence of purely psychological avoidance, a cutting off of one's normal curiosity by an effort of will" is enough. Stone, 987 F.2d at 472; United States v. Giovannetti, 919 F.2d 1223, 1229 (7th Cir.1990).

Several circumstances in this case support such an inference. Briefly, we mention the following: Sanchez's patently false letters and claims of authority, which Dumlao never inquired about, despite warnings from her receptionist; the "to whom it may concern" notes, which contained promises not to sell medication but were never followed up on or enforced in any way; Dumlao's failure to return calls from IDASA warning her about professional patients, while calling officials there to obtain new script pads; Dumlao's failure to conduct meaningful tests to ascertain her patients' true medical problems; her failure to inquire into the whereabouts of her patients' medical records; her failure to question why she was the only doctor in the entire world who could supply these patients with medication, etc. We have also reviewed Dumlao's testimony at trial and, like the trial judge, find her statements questionable and unresponsive at several points. There may be a fine line between "deliberate ignorance" and "should have known," but viewing this evidence in the light most favorable to the government, that line was clearly crossed in this case. Judge Nordberg's carefully considered decision to include the ostrich instruction was appropriate and well within his discretion.

Dumlao next claims that the trial court violated her Fifth Amendment right to due process because it refused to instruct the jury on her entrapment defense. We review a decision whether to give an entrapment instruction de novo. United States v. Santiago-Godinez, 12 F.3d 722, 726 (7th Cir.1993), cert. denied, --- U.S. ----, 114 S. Ct. 1630, 128 L. Ed. 2d 354 (1994).

The defense of entrapment requires proof of two elements: (1) government inducement of the crime, and (2) a lack of predisposition on the part of the defendant to engage in the criminal conduct. Mathews v. United States, 485 U.S. 58, 63, 108 S. Ct. 883, 886, 99 L. Ed. 2d 54 (1988); Santiago-Godinez, 12 F.3d at 728. The first question here is whether there was sufficient government involvement, since there is no defense of private entrapment. United States v. Manzella, 791 F.2d 1263, 1269 (7th Cir.1986). Dumlao argues that because of Sanchez's contacts with the government as an informant, which continued during the period of the conspiracy, Sanchez should be considered a government agent for purposes of the entrapment defense. She asserts that the DEA had full knowledge of Sanchez's conduct in relation to her activity with Dumlao, and that in exchange for information concerning illegal drug dealing, Sanchez received an "implied license" to continue her own illegal Dilaudid dealings.

The district court's findings on this issue were quite different, and we concur in that judgment. Sanchez became a government informant in March of 1983, but was deactivated in March 1986. During this three-year period she was paid a total of only $1200 for her information. She also agreed to certain terms in this relationship included in a signed statement as follows:

I will not violate any criminal laws in the gathering of my information or evidence. I understand that if I do violate any criminal laws, I will be prosecuted. I will not receive or possess any narcotics without the express approval and the presence of a Drug Enforcement Agent.

After 1986 Sanchez was considered only a "source of information," and spoke to officials occasionally on a voluntary basis. She gave officials limited information on certain individuals, but minimized or omitted her own role in drug transactions. Significantly, she never informed any government representative of her contacts with Dumlao, and even took affirmative steps to hide these, advising one of her professional patients to not tell the government about "the new doctor." In sum, during her entire four-year association with Dumlao, Sanchez kept the prescriptions, the pills, and the proceeds related to her Dilaudid dealings, never turning anything over to the government. Apparently she sought not to help the government, but to limit her competition, shield her scheme, and/or give up her friends, if necessary, to protect herself. That does not suggest an agency relationship with the government. See United States v. Busby, 780 F.2d 804, 806-07 (9th Cir.1986) (holding that previous activities as paid informant do not establish an agency relationship). The district court's refusal to tender the entrapment instruction was fully justified on the basis of government involvement alone, thus we need not consider the further questions of inducement or predisposition.

Related in part to Dumlao's claim of entrapment, Sanchez contends that the district court erred in ruling that she could not present a defense of "public authority," sometimes called "entrapment by estoppel." In general terms this defense permits an acquittal when the defendant was reasonably mistaken in believing her criminal activity was authorized by the government. Sanchez claims that since this was a theory of her defense, the district court's refusal to give the instruction effectively deprived her of a fair trial in violation of the Fifth Amendment.

Defendants are entitled to a theory of defense instruction if: (1) they propose a correct statement of the law; (2) their theory is supported by the evidence; (3) their theory is not part of the charge; and (4) the failure to include an instruction on the defendant's theory of defense would deny him or her a fair trial. United States v. Howell, 37 F.3d 1197, 1203 (7th Cir.1994), cert. denied, --- U.S. ----, 115 S. Ct. 1810, 131 L. Ed. 2d 735 (1995); United States v. Elder, 16 F.3d 733, 738 (7th Cir.1994). In reviewing these matters, "[w]e defer to the substantial discretion of the district court for the specific wording of the instructions and for the decision to reject a proposed instruction, as long as the essential points are covered by the instructions given." Howell, 37 F.3d at 1203-04; United States v. Scott, 19 F.3d 1238, 1245 (7th Cir.), cert. denied, --- U.S. ----, 115 S. Ct. 163, 130 L. Ed. 2d 101 (1994).

The defense of entrapment by estoppel is rarely available. Howell, 37 F.3d at 1204. "In essence, it applies when, acting with actual or apparent authority, a government official affirmatively assures the defendant that certain conduct is legal and the defendant reasonably believes that official." Id. To employ this defense, we have required that the government official "actively mislead the defendant; and that the defendant's reliance be actual and reasonable in light of the identity of the agent, the point of law represented, and the substance of the misrepresentation." Id. (footnote with extensive case citations omitted).

The district court found that the evidence was insufficient to support this defense--Sanchez had received no authorization from the government to commit the crimes of which she was convicted, and she could not have reasonably believed that she was acting pursuant to official authorization during the conspiracy. The district court did, however, allow Sanchez to put on evidence of the extent of her cooperation with law enforcement officers; she called seven federal law enforcement agents to testify concerning their relationship with her. Sanchez's counsel also reminded the jury in his closing argument of this theory, concluding his presentation by asking, "If you're going to play Frankenstein, do you take responsibility for the monster?"

The district judge was correct in concluding that any monster here was of Sanchez's own making. The only "authorization" Sanchez mentions in her brief is the claim that DEA agent Paul Galvin "directed Sanchez to produce prosecution witnesses" in the following conversation:

Galvin: You know what would really help them? If, if, instead of waiting for us to come out and see them, if they really wanted to show that they wanted to cooperate.

Sanchez: I can get Carol [Turner] and them to come down.

Galvin: You should get, yes, get Carol and them to come down. You know where our office is. If they come down on their own free will and tell us the truth, it'll, it'll really help them out.

We fail to see how this constitutes an authorization to "produce witnesses," especially since they must come of their own free will, let alone how it constitutes an authorization to commit the many other crimes involved here. We also fail to see anywhere in the record any "active misrepresentation" which authorized Sanchez to obtain, deal, and profit from Dilaudid to the tune of hundreds of thousands of dollars. Her actions within the Dilaudid conspiracy make her claim of public authority palpably incredible; the relevant evidence of her minimal and opportunistic interaction with the DEA has been recounted above. Even if Sanchez somehow truly believed that she was indeed a "government agent," this is unavailing, for reasonableness in this context is objective. United States v. Barker, 546 F.2d 940, 947-48 (D.C.Cir.1976). And since the district court allowed the relevant aspects of her theory to be introduced through the evidence and closing arguments, Sanchez was not denied a fair trial. None of the requirements necessary to warrant the instruction were met here, and it was proper to exclude it.

Neville and Sanchez argue that their rights under the Speedy Trial Act, 18 U.S.C. § 3161 et seq., were violated by the delay between the discharge of the first jury seated to hear their trial on December 4, 1992, and the beginning of the trial on April 7, 1993. The defendants admit much of this four-month period was excludable under the exceptions provided in the Speedy Trial Act, see 18 U.S.C. § 3161(h), but claim the district judge's use of the "ends of justice" exception, see 18 U.S.C. § 3161(h)(8), to justify the remainder of the delay was error.

The Speedy Trial Act mandates that a trial of any case in which a plea of not guilty is entered shall begin within seventy days of the information or indictment, or from the defendant's appearance before a judicial officer, whichever date last occurs. 18 U.S.C. § 3161(c)(1). The statute provides several exclusions from the calculation of this time limit, including any delay where the court finds "that the ends of justice served by the taking of such action outweigh the best interest of the public and the defendant in a speedy trial." 18 U.S.C. § 3161(h)(8)(A). If a court relies on this "ends of justice" rationale, reasons for the delay must be set forth either orally or in writing. Id. We review the district court's justifications for abuse of discretion noting that, absent legal error, exclusions of time cannot be reversed except where there is abuse and a showing of actual prejudice. See United States v. Scott, 784 F.2d 787, 789 (7th Cir.), cert. denied, 476 U.S. 1145, 106 S. Ct. 2257, 90 L. Ed. 2d 702 (1986).

Jury selection for this trial began on November 23, 1992. On November 25, the district court ordered a competency examination for codefendant Louis Lacour. On the basis of that exam, on December 4 the court found Lacour unfit to stand trial.4 Trial was thus continued until April 5, 1993, and began shortly thereafter. The district court found the delay between November 20, 1992 and April 5, 1993 excludable under 18 U.S.C. §§ 3161(h)(3)(A) and (B) (absence or unavailability of defendant), (h)(7) (codefendant's time for trial has not run, and no severance has been granted), and (h)(8)(B)(i-iv) ("interests of justice" factors).

We find these justifications accurate and sufficient. Because of ongoing competency investigations, Lacour, a key member of this large and complicated conspiracy, was "unavailable" for trial under either § 3161(h)(3)(A) and (B), or (though not relied on) § 3161(h)(4) (mental incompetence or physical inability). Furthermore, we have held that under § 3161(h)(7), applicable here, the excludable delay of one defendant may be ascribed to all codefendants, if the delay is reasonable. United States v. Dennis, 737 F.2d 617, 620-21 (7th Cir.), cert. denied, 469 U.S. 868, 105 S. Ct. 215, 83 L. Ed. 2d 145 (1984). Considering the purpose and duration of the delay here, it was not unreasonable, and thus the time attributable to Lacour is excluded as to Sanchez and Neville also.

The trial court also cited the "interests of justice" exclusion in order to attempt to try all of the defendants together and avoid the need for a separate trial. We have recognized "a strong congressional preference for joint trials and an intention that delays resulting from the joinder of codefendants be liberally construed," Dennis, 737 F.2d at 621, and thus find no abuse of discretion in this rationale. The defendants do not claim prejudice as a result of the delay--only 120 days total--nor do they challenge on appeal the district court's failure to grant a severance. We find no merit in their claims of undue delay.

Finally, all defendants challenge the calculations of their offense levels for sentencing purposes. Essentially, they claim that the court should have used the weight of the hydromorphone in each tablet rather than the weight of the entire tablet in arriving at its conclusion, and base this theory on the fact that the "Drug Equivalency Table" in Sentencing Guidelines § 2D1.1 uses the term "hydromorphone" rather than "Dilaudid."

Defendants acknowledge that we have addressed this overall issue previously in United States v. Lacour, 32 F.3d 1157, 1159-61 (7th Cir.1994). That decision, like that of every other court of appeals to address the issue,5 held that the sentence of a defendant convicted of an offense involving Dilaudid should be calculated "based upon the gross weight of Dilaudid tablets." Id. at 1160. That case also considered defendants' specific argument here, and rejected it on the basis of a note to the Drug Quantity Table which states that "[u]nless otherwise specified, the weight of a controlled substance set forth in the table refers to the entire weight of any mixture or substance containing a detectable amount of the controlled substance." Id. (citing U.S.S.G. § 2D1.1(c) n.*). We see no reason to revisit our previous holding.

For the foregoing reasons, the convictions of the defendants are AFFIRMED.