Visual Identification of Drugs (Again)

linkedin
Share on Google+
Share on Reddit
Share on Tumblr
Download PDF

The longest opinion issued by the court of appeals this week was Judge Ervin’s 45-page treatise in State v. Ward, __ N.C. App. __ (2009). Although the opinion contains other important material, I want to focus on the court’s holding that the method used by an SBI agent to identify certain prescription drugs was “not consistent with the general thrust of existing precedent concerning how controlled substances should be identified in criminal trials and . . . [was] not sufficiently reliable.”

The defendant in Ward sold what appeared to be prescription drugs to a police informant. A subsequent search of the defendant, his car, and his house turned up enough additional drugs to stock a pharmacy. He was charged with, and convicted of, a variety of drug offenses. Part of the state’s evidence at trial was the testimony of an SBI agent who testified as an expert in the fields of chemical analysis of drugs and forensic chemistry. He testified that “that he performed a chemical analysis or visual examination of the evidence seized from Defendant . . . [and that] these substances included Cocaine, Dihydrocodeinone (an opium derivative), Hydrocodone (an opium derivative), Oxycodone (an opium derivative), [and various other controlled substances].” He indicated that he identified some of the drugs based upon a chemical analysis, but that he “identified certain other substances on the basis of a visual examination of the size, shape, color of and markings on the tablets in question.” Specifically, the agent compared the appearance of the pills, including their pharmaceutical markings, to a reference work called Micromedics Literature, which he testified is “used by the doctors in hospitals and pharmacies to identify prescription medicine[s].”

On appeal, the defendant argued that the trial court erred in allowing this testimony, claiming that the method used by the agent to identify the drugs was not sufficiently reliable to satisfy the requirements of Rule 702 and State v. Goode, 341 N.C. 513 (1995). The court of appeals agreed. First, it addressed State v. Fletcher, 92 N.C. App. 50 (1988) (allowing a police officer to testify as an expert and to identify marijuana based on its appearance), discounting it as a pre-Goode case that was not focused on the reliability of the testimony at issue. The court then noted without comment that Fletcher was applied (post-Goode) to crack cocaine in State v. Freeman, 185 N.C. App. 408 (2007) (allowing a police officer to give his lay opinion that a particular substance was crack cocaine), before focussing on State v. Llamas-Hernandez, 363 N.C. 8 (2009), reversing for reasons stated in dissenting opinion, 189 N.C. App. 640 (2008), a case discussed at length in this prior blog post.

Llamas-Hernandez held that two police officers should not have been allowed to give lay opinion testimony to the effect that a white powder seized during an investigation was cocaine. The court noted that powder cocaine does not possess unique visual characteristics; that the General Assembly adopted “a technical, scientific definition of cocaine,” suggesting the need for scientific testimony to identify it; and that the General Assembly provided a statutory procedure for the admission of lab reports regarding controlled substances, perhaps suggesting that laboratory analysis is required. Because Llamas-Hernandez did not overrule Freeman and Fletcher, the scope of the ruling was somewhat unclear: was it limited to powder cocaine, or did its reasoning sweep more broadly?

Ward applied Llamas-Hernandez to prescription drugs, stating that “existing precedent suggests that controlled substances defined in terms of their chemical composition can only be identified through the use of a chemical analysis rather than through the use of lay testimony based on visual inspection.” It also noted that the agent had received no special training in identifying prescription drugs, and that there was little reason to believe that the agent could reliably distinguish between genuine and counterfeit prescription drugs, which the court, based on World Health Organization data, believed to be widespread. Thus, it concluded that the agent’s testimony was insufficiently reliable to be admissible.

A few minutes’ worth of research suggests that most other jurisdictions have come out on the other side of this issue. See, e.g.State v. Clark, 198 P.3d 809 (Mont. 2008) (allowing pharmacist to testify to the identity of prescription medications “by comparing the unique imprint code on each pill, along with the color, make, and shape of the pills, with two national peer-reviewed computer databases”); State v. Carter, 981 So.2d 734 (La. Ct. App. 2008) (finding sufficient evidence to support a conviction where “an expert in forensic chemistry . . . testified that the green pills found by police contained hydrocodone. The pills were identified via visual inspection and comparison with pictures in a book.”); State v. Ohlin, 2007 WL 4106274 (Minn. Ct. App. Nov. 20, 2007) (unpublished) (“[A] forensic scientist . . . inspected the pills and identified them from their markings as containing a controlled substance. There is nothing in the record to contradict this expert testimony or to support a plausible inference that the pills contained anything different from what their markings showed. Nor has [the defendant] shown why it would be improper for an expert witness to rely on pill markings, in lieu of testing, to determine the presence of a controlled substance in a pill found in a prescription bottle and not packaged for sale.”). Of course, that doesn’t show that Ward was wrongly decided, but it does suggest that this area remains controversial. Perhaps the state supreme court will weigh in and clarify the reach of Llamas-Hernandez. Stay tuned for further developments.

5 comments on “Visual Identification of Drugs (Again)

  1. Where does this leave us with lay identification of marijuana?

  2. Ward doesn’t overrule Fletcher, nor does Llamas-Hernandez, so if I were a trial judge I would feel bound to admit testimony like that in Fletcher. (Technically, that was expert testimony, not lay testimony, though in light of Freeman, I’m not sure that’s a critical distinction.) But if I were a prosecutor, I would be a little nervous about my prospects on appeal absent a chemical analysis.

  3. I just want to make sure I understand the underlying factual issue here. Did the SBI lab analyst identify the pills only by sight, or did the agent chemically test a sample of the pills to verify what they were, then use the chemical analysis on the sample to give an opinion about the whole batch?

    If it is the former, then I understand the opinion. If it is the latter, than there will be a huge issue because then the SBI will have to chemically test every single pill to verify what it is.

  4. It’s the former. The opinion doesn’t appear to affect prior case law permitting sampling.

  5. I just want to make sure I understand the underlying factual issue here. Did the SBI lab analyst identify the pills only by sight, or did the agent chemically test a sample of the pills to verify what they were, then use the chemical analysis on the sample to give an opinion about the whole batch?

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.