Though Texas has had a breath alcohol testing program for over 40 years, its program has only recently adopted formal written “Standard Operating Guidelines” (SOG) for technical supervisors to follow in the operation of some aspects of the breath test program. The lack of such guidelines (or almost any written policies) has long been problematic and has long been a source of valid criticism of the Texas program—from within Texas and elsewhere. Why it has taken this long to begin to try put anything resembling science into a program headed by a position named “Scientific Director” is, at least in part, due to the culture at DPS that believed that if it had written policies and standards, others might seek to hold those involved in the program accountable to the standards, both scientifically and legally. One can speculate why DPS would see the need to be concerned about those in the program living up to specified standards and policies, but, at least until they begin to water them down, the Texas Breath Alcohol Testing Program now has written standards that technical supervisors are required to follow with respect to some aspects of the breath testing program. It is a new paradigm in Texas—and one this article will explore, both from scientific and legal perspectives.
The new SOGs have fundamentally changed the way DPS regulates breath alcohol testing in Texas. Rather than regulating breath alcohol testing exclusively through formal adoption of administrative regulations in the Texas Administrative Code, DPS now also regulates breath testing through rules contained in its “Standard Operating Guidelines,”which the administrative regulations require to be followed both to have a program approved and to maintain continued approval and certification of the program.
DPS has long had the ability to promulgate and adopt administrative regulations governing breath alcohol testing.1 Such regulations have been and are contained in Title 37, Part 1, Chapter 19, Subchapter A of the Texas Administrative Code. 37 TAC § 19.1 et seq. Until recently, such administrative regulations were the beginning and end of formal policies by DPS concerning breath testing—at least so far as technical supervisors would admit.2 Instead of having requirements exclusively in the Texas Administrative Code, 37 TAC § 19.4(f) of the administrative regulations provides that approval of breath alcohol testing programs is contingent on each program agreeing “to conform and abide by any directives, orders, or policies issued or to be issued by the scientific director regarding any aspect of the breath alcohol testing program.”3 Thus, the administrative regulations incorporate a requirement that technical supervisors and breath test operators also comply with “any directives, orders, or policies issued or to be issued by the scientific director. . . .”
Most judges, prosecutors and defense lawyers are familiar with the requirement that to be admissible, a breath test cannot violate and must comply with the requirements of the Texas Administrative Code.4 In large part, this knowledge has been the byproduct of the administrative regulations being the only formal rules applicable to breath testing. Now, however, the SOGs, as well as other rules and directives of the department, govern the administration of breath alcohol tests in Texas.
Why the SOGs Matter to Breath Test Cases: Suppression for Failure to Follow
This article is, in part, the product of participants in the criminal justice system wanting to know more about the new SOGs, and about them asking whether a breath test result is subject to suppression if it does not comply with the SOGs. In answering this question it is necessary to recall that it is often amazing what one finds when actually reading statutes and other laws. Tex. Trans. Code § 724.016 provides:
(a) A breath specimen taken at the request or order of a peace officer must be taken and analyzed under rules of the department by an individual possessing a certificate issued by the department certifying that the individual is qualified to perform the analysis.
(b) The department may:
(1) adopt rules approving satisfactory analytical methods.5
Section 724.016 has long been held to set a statutory predicate for admissibility of breath tests in Texas. That is, a breath test must comply with § 724.016 before it is admissible, and a breath test that was not “taken and analyzed under the rules of the department” and that violated the rules of the department was not admissible.6
Notably, § 724.016 does not limit its applicability to regulations adopted under the Texas Administrative Code. It applies to all “rules” of the Department. The new SOGs constitute “rules” of the department. First, on their face, they are rules that must be followed by technical supervisors in administering some aspects of a breath alcohol testing program. In the common understanding of the word “rules,” the SOGs are rules of the department that must be followed.
Second, and likely more importantly, § 19.4(f) of the administrative regulations expressly requires compliance with “any directives, orders, or policies issued or to be issued by the scientific director regarding any aspect of the breath alcohol testing program.” The SOGs clearly constitute “directives, orders, or policies issued . . . by the scientific director. . . .” Indeed, under § 19.4(g), approval of the program may be withdrawn if a program does not comply with all requirements of § 19.4, which includes the requirement to abide by “any directives, orders, or policies issued or to be issued by the scientific director regarding any aspect of the breath alcohol testing program.”7 In this context, the SOGs expressly refer to the contents of the SOGs as “policies” of the department. Thus, regardless of whether the SOGs are independently “rules” within the meaning of § 724.016, they are incorporated into § 19.4 of the administrative regulations which have long been held to be within the scope of § 724.016.
Amend v. State, No. 05-08-01656-CR, 2010 Tex. App. LEXIS 3315 (Tex. App.—Dallas, Mar. 5, 2010, no pet.)(not designated for publication), presented a factual situation where what constituted a “rule” of the department under § 724.016 and whether an oral directive from the Scientific Director constituted “any directives, orders, or policies issued or to be issued by the scientific director” under § 19.4(f). In Amend, the defendant argued that an oral directive from the Scientific Director that the 15 minutes presence period contained in 37 TAC § 19.4(c)(1) was not satisfied during the time that the observing officer was in the front seat of a patrol car and the breath test subject was in the back seat, at least when they subsequently arrived at the station and the defendant was taken inside for a later breath test. Though the court ultimately resolved the issues on the basis that the oral directive was not in effect at the time of Amend’s breath test, the factual scenario repeats itself throughout the state on a regular basis and technical supervisors other than the one in Amend have given substantially similar testimony.8
There is no reason to limit the applicability of § 724.016 to only written “rules” when the statute contains no such limitation. Similarly, nothing in § 19.4(f) limits the “directives, orders, or policies . . . of the scientific directive” only to those issued in writing. Indeed, it is more likely that the use of the word “any” before “directives, orders, or policies” without any limiting language makes it clear that oral “directives, orders, or policies” are included. This would also be consistent with the long established DPS practice of often issuing “directives, orders, or policies” only in oral form.
The scope of § 724.016 and § 19.4(f) also makes it important for counsel to seek from the technical supervisor, via discovery in each case, all rules, directives, orders, and/or policies, other than anything contained in the Texas Administrative Code:
1. applicable at the time of the breath test in the case;
2. at the time of certification of the breath test machine in the case;
3. at the time of the last calibration, as defined by the SOGs, of the breath test machine in the case;
4. at the time of the last “Autocal”9 of the breath test machine in the case; and
5. at the time of the preparation of all solutions used to Autocal, calibrate, or in the reference simulator used at the time of the test in the case.
While some technical supervisors have testified that there are no rules, directives, orders, and/or policies other than what is contained in the Texas Administrative Code, other technical supervisors have testified differently, and some technical supervisors have produced documents purporting to be rules, directives, orders, and/or policies issued by the Scientific Director. Obviously, either not all are telling the truth or not all technical supervisors are aware of all of the rules, directives, orders, and/or policies issued by the Scientific Director: Either alternative has potentially adverse consequences for the State in the trial of a breath test case.
The minimization and elimination of details and specifics in § 19.4 of the administrative regulations was likely the product of a desire to have more flexibility in modifying breath alcohol testing program rules without having to go through the cumbersome and formal procedures of amending the administrative regulations. Without doubt, it has accomplished that goal. The changes have also made the requirements of the rules in the SOGs predicates to admissibility and a basis for suppression when they have been violated.
The elements of the rules, regulations, directives, order, and policies may only become part of the predicate for admissibility when an issue about a specific rule is expressly raised by the defense. In Gifford v. State, 793 S.W.2d 48, 49 (Tex. App.—Dallas 1990), pet dism’d improv granted, 810 S.W.2d 225 (Tex. Crim. App. 1991), the Court, relying on Harrell v. State, 725 S.W.2d 208 (Tex. Crim. App. 1986), held that compliance with the 15-minute observation period was not part of the predicate for admissibility of the Intoxilyzer test result. Further, relying on Sims v. State, 735 S.W.2d 913 (Tex. App.—Dallas 1987, pet. ref’d), and Mosely v. State, 696 S.W.2d 934 (Tex. App.—Dallas 1985, pet. ref’d), the Court further held that the State was only required to prove compliance with DPS regulations when a defendant had “raised a fact issue” as to whether there had been compliance with a particular regulation.
Regrettably, the cases on which Gifford relied do not stand for the propositions for which they were cited and relied upon by the court. There is not now and never has been a rule in Texas that a statutorily required predicate is satisfied merely because a defendant does not offer evidence that it has not been satisfied, i.e., by creating a fact issue. Harrell does not expressly hold that the State need not show compliance with the 15-minute continuous observation period in order to admit an Intoxilyzer test result. Harrell does, however, hold that the predicate for admissibility of a breath test result is: (1) the use of properly compounded chemicals; (2) the existence of periodic supervision over the machine and operation of the machine by one who understands the scientific theory of the machine; and (3) proof of the result of the test by a witness qualified to translate and interpret such result so as to eliminate hearsay. Despite its apparent omission of the 15-minute observation period, the lack of discussion of it is easily explained.
First, the observation period was not a regulatory requirement when the three-prong test was first promulgated. See French v. State, 484 S.W.2d 716 (Tex. Crim. App. 1972); Hill v. State, 256 S.W.2d 93 (Tex. Crim. App. 1953). Since the 15-minute observation issue was not at issue in Harrell, it is neither surprising nor of any legal significance that the Court did not modify the long-standing three-part test. Second, and more importantly, the only issue in Harrell was whether the breath test at issue complied with DPS regulations. Thus, regardless of whether the compliance with the regulatory requirements was part of the formal three-part test, the Court implicitly recognized that compliance with DPS regulations was necessary for a breath test result to be admissible. Otherwise, the entire opinion was meaningless dicta.
Ultimately, the court held that the breath test was admissible because it complied with DPS regulations. Thus, while proof of the 15-minute observation period may not be part of the three-part test initially discussed in Harrell, it is not accurate to say that a breath test is admissible without regard to proof of compliance with DPS regulations. To the contrary, and consistent with the statute, the State must show that the breath test was conducted in accordance with applicable DPS regulations.
The Court’s reliance on Mosely is similarly misplaced and erroneous. While Mosely did hold that a defendant is not entitled to a jury instruction unless the evidence has raised a fact issue, Mosely did not hold that a fact issue was necessary before the State had to show compliance with DPS regulations for admissibility purposes. To the contrary, the court held that for admissibility purposes, the State had showed compliance with all applicable DPS regulations by virtue of undisputed, unobjected-to evidence of such compliance. Mosely, 696 S.W.2d at 936. Indeed, the court in Sims, 735 S.W.2d at 919, the final case erroneously relied on by the court in Gifford, expressly recognized that the “fact-issue” holding of Mosely applied only to whether a defendant was entitled to a jury instruction on compliance with DPS regulations.
What Gifford may have meant, but did not say, is that proof of compliance with every detailed regulation is not necessary in the absence of a specific objection pointing out the alleged failure to comply with the regulations. Indeed, this is suggested by Mosely. 696 S.W.2d at 936–937. Finally, as Mosely points out, the Court of Criminal Appeals has held that proper administration of the test, i.e., compliance with DPS regulatory requirements for administering the test, is “a necessary predicate to admissibility of the test result when lack of such a predicate is raised by a proper objection.” Mosely, 696 S.W.2d at 936. See Slagle v. State, 570 S.W.2d 916, 917–918 (Tex. Crim. App. 1978); Cody v. State, 548 S.W.2d 401, 404 (Tex. Crim. App. 1977). It is an objection that raises the prospect of the need for additional proof, not the creation of a fact issue, when admissibility is the issue.
Hawkins v. State, 865 S.W.2d 97 (Tex. App.—Corpus Christi 1993), also holds that the State was not required to prove compliance with the 15-minute observation period contained in the DPS regulations when the defendant had not “raised a fact issue” as to whether there had been compliance with the 15-minute observation requirement. Hawkins relied exclusively on the unsupported holding of Gifford. Thus, Hawkins is likewise suspect. Moreover, it appears that what really happened in Hawkins is that the defendant never objected to the lack of proof of the observation period. The court’s opinion couched the evidence in terms of Hawkins never having “averred” that the trooper did not follow the 15-minute observation period. Once again, it is the specificity of an objection that raises the issue, not the need for additional evidence to create a fact issue.
Counsel handling breath test cases need to ensure that they have a thorough knowledge of the “rules” as well as the “directives, orders, and/or policies” in effect at the time of various steps in the breath testing process. Once counsel knows the standards applicable to the case, counsel can determine whether the State can show compliance with the requisite standards. If the State cannot prove compliance, either because proof is not available to them or because there was not compliance, suppression of the breath test and a favorable resolution of the case become far more likely.
The SOGs: Scope, Details, and Opportunities
The SOGs contain six parts.
- Part One: “Instrument and Testing Location Management”
- Part Two: “Record Keeping: Instrument Certification File, Solution File, Subject File, and Maintenance File”
- Part Three: “Electronic Data Management”
- Part Four: “Record Retention and Release”
- Part Five: “Breath Alcohol Instrument Calibration”
- Part Six: “Thermometers Used in the Instrument Calibration Procedure.”
Part One: “Instrument and Testing Location Management”
This section governs the procedures that must be followed by the Technical Supervisor (TS) in inspecting and testing the machine. It describes what constitutes a “complete inspection.” Section 1.1 provides that an inspection of an evidential machine can only be conducted at the testing site. Thus, it would be improper to inspect and test a machine at any location other than the “evidential testing location.” Issues with respect to this requirement could arise when a machine that is normally at one location is moved to another location or to a mobile location or when a TS purports to have conducted an inspection at home or in a lab followed by a later transfer of the machine to a testing location.10
Section 1.1.1 states a specific requirement for a proper inspection: The “capability of the instrument to detect and subtract the effect of acetone shall be tested.” Though the rule requires the TS to test both the ability of the machine to detect and subtract acetone, subsection 1.1.1.1 provides a specific procedure only for detecting acetone. No procedure is specified for inspecting the “capability of the instrument to . . . subtract the effect of acetone. . . .” Practitioners will be unlikely to find any support for a claim that the machine’s ability to subtract acetone was tested. If this portion of the rules has not been followed, a breath test conducted after such a deficient inspection is not in compliance with the rules of the department and the requirement in Harrell for periodic inspection (in compliance with DPS rules) has not been satisfied. Counsel should seek discovery of documentation showing compliance with this rule.
Section 1.2 requires a complete inspection each time a machine is placed into or returned to service at a location. Section 1.2.1 requires that each time a machine is placed into or returned to service at a location the TS shall hand write on the INSP TS test record the date on which the machine was last calibrated and autocaled. Section 1.3 requires that if possible, a compete inspection shall be done each time a machine is removed from service at a testing location. Section 1.4 requires that every machine have a complete inspection at least once each calendar month.
Section 1.5 requires that at least once each calendar month, the reference solution should be replaced with a new solution and the lot number of the new solution recorded on the test record. Section 1.5.1 requires that the nominal value (the expected value) of the reference solution shall be set as the “predicted value.” Thus, if the TS thinks they have a 0.08 solution, 0.08 shall be set as the predicted value. Importantly, section 1.5.1.1 requires that the nominal value be within 0.003 or 3 percent, whichever is greater, of the unbiased estimate of the alcohol concentration.11 Thus, a 0.08 nominal solution value must actually be no less than 0.077 and no more than 0.083.
Most importantly, section 1.5.2 requires that the unbiased estimate and the uncertainty (plus or minus a specified amount) of the solution shall be established by an unspecified “procedure establishing traceability of measurement through an unbroken chain of comparisons to the National Institute of Standards and Technology (NIST) all having stated uncertainties.” Traceability to NIST is a common scientific requirement for assuring accurate measurement. NIST is a government organization that established the true values of various weights and measures. Think of it this way: What is really one gram? NIST answers that question by providing weights that are certified to be one gram plus or minus some degree of uncertainty. Traceability to NIST is a process that requires whatever is being measured to be compared against a known value from NIST. Thus, if a TS uses a solution prepared for the reference simulator solution, that solution must have been tested in comparison to a known standard from NIST. Often times, commercially prepared solutions represent their reported value (plus or minus uncertainty) to be traceable to (measured against) a NIST standard.
The requirement for NIST traceability in breath testing is new in Texas. Counsel should always seek discovery of documentation verifying NIST traceability. As a practical matter, a TS in Texas may have problems testifying to NIST traceability unless that TS has actually done the comparison. A TS who seeks to testify to traceability merely because someone else has told him that it is traceable may be offering no more than hearsay. Such testimony may also raise confrontation issues.
Part Two: “Record Keeping: Instrument Certification File, Solution File, Subject File, and Maintenance File”
This section contains requirements for documenting the certification and calibration of the machine. Technical Supervisors are required by section 2.1 to maintain an “Instrument Certification File,” which must contain “Evidential Breath Alcohol Testing Instrument Certificates” and “Letters of Certification.” Section 2.1.1 requires that prior to being placed into service an instrument must have either a current “Calibration Certificate” and an “Evidential Breath Alcohol Testing Instrument Certificate” or an instrument certification letter signed by the Scientific Director. Section 2.1.2 explains that when a TS provides the Office of the Scientific Director with an initial Calibration Certificate, the Scientific Director will issue an Evidential Breath Alcohol Testing Instrument Certificate. Though it might seem like these paperwork requirements would be routinely followed, hundreds of breath alcohol results in Polk County were invalidated a few years ago when similar documentation requirements had not been fulfilled. Counsel should always seek discovery of this documentation.
Section 2.2 provides that the “Solution File” maintained by the TS must contain the certificates of Analysis and documents necessary to establish NIST traceability, unbiased estimate, nominal value, and combined uncertainty of the reference sample solution used at evidential testing locations. Counsel should always seek discovery of this documentation.
Section 2.3 provides that a TS shall collect and maintain a copy of each breath alcohol test record produced by a machine in the “Subject File.” Section 2.3.1 provides that evidential subject test records may only be produced by a certified instrument in an approved evidential testing location, and that such records must be sequentially numbered. Section 2.3.2 provides that no alterations may be made to data on a printed subject test record. Thus, if the breath test operator has erred in entering information for a breath test, that data may not later be altered. Section 2.3.3 provides a procedure when a breath test slip is missing. Section 2.3.4 provides that test records produced in a laboratory are not evidential and must not be kept in the Subject File. Discovery requests should seek both test slips in the Subject File as well as any non-evidential slips not kept in the Subject File.
Section 2.4 governs records in the “Maintenance File” pertaining to instrument and simulator maintenance and repair and certificates of calibration for thermometers used in the calibration procedures. Calibration procedures are discussed later. Section 2.4.1 provides that maintenance records shall be organized and detailed enough to allow any other TS to render an opinion in litigation concerning the maintenance history of the machine. Thus, it is not enough that a TS claims to “know” what he or she has done on and to the machine; others must be able to also “know” it just from looking at the records. Section 2.4.2 requires that each INSP TS (see section1.1.2) record be kept in the Maintenance File. Section 2.4.3 requires that each ACA test record generated by an inspection “to demonstrate the ability the instrument to detect and subtract acetone,” as required by section 1.1.1.1, be kept in the Maintenance File.12
Section 2.4.4 provides that the Maintenance File shall contain notes “regarding the major components . . . that are replaced or repaired.” Section 2.4.4.1 provides that if abbreviations or codes are used to describe maintenance and repairs, there shall be a document clearly explaining the codes and abbreviations. Section 2.4.5 requires that the Maintenance File contain notes of major components of each simulator that is replaced or repaired. Section 2.4.6 requires that the Maintenance File contain documentation establishing the NIST traceability of thermometers used in the calibration procedures.
Every document in the Maintenance File should be sought in discovery. The necessary time frame for each type of document will vary from category to category. For example, the necessary time frame for records of major repairs should be substantially longer (perhaps even years) than the INSP TS and acetone ACA records.
Part Three: “Electronic Data Management”
DPS has begun taking a new position with respect to reports generated from its electronic data. For decades, TS’s provided and defendants regularly obtained reports created from the electronic data. These reports were basically printouts of spreadsheet data and included such useful information as actual measured values when the reference solution was out of tolerance and error codes for any test in which there was an error. For decades, TS’s preached that defendants did not need individual breath test slips because TS’s preferred just to provide the breath test log generated from the electronic data. These logs allowed voluminous data to be viewed in a summary fashion that made it more useful and user friendly. DPS has now done an about-face and is instructing TS’s not to produce the breath test logs that they have long represented was all the defendants really needed in discovery. The new desire to only provide test slips is nothing more than an effort to limit information dissemination that could be useful in the defense of a breath test case.13
Section 3.1 sets out how useful the electronic data can be. Section 3.1.2 governs when a TS may manually enter data into the database in place of electronic data. The ability to manually enter data is part of what allowed Dee Wallace (a former technical supervisor) to fabricate on-site inspection records.14 Following this debacle, DPS required specific and explicit permission before any TS could manually enter or change data. Regrettably, the SOGs have eliminated the protections put in place to prevent the manual fabrication of records.
Section 3.2 governs TS monthly reports. Section 3.2.1 and its subsections govern the processing of complete and incomplete monthly TS reports. Counsel should seek discovery of these reports—both complete and incomplete.
Part Four: “Record Retention and Release”
Part Four governs record retention and release. Section 4.1 provides that test records produced at evidential testing location shall be kept for five years plus the current year. Section 4.2 provides that records “shall be provided by the Technical Supervisor when requested through open records requests, discovery motions, and subpoenas.” Section 4.2.1 provides that “any document held in the possession of the technical supervisor is considered to be a government document as defined by Texas Government Code § 552.002,” and that records can be either electronic or paper. Texas Penal Code § 37.10 makes it a felony when a person “intentionally destroys . . . a government record.” Counsel should investigate whether a crime has occurred any time a TS destroys the records created by the Intoxilyzer each time an “Autocal” is performed on the machine. (See the discussion below regarding calibration and “Autocal” records.)
Section 4.3 provides that “[e]lectronic data is important . . .” Section 4.3.1 seeks to implement DPS’s change of heart regarding breath test logs. In doing so, however, it established that there is meaningful data in the electronic records that does not appear on breath test slips. Although this section “recommends” that only printed test records be submitted to comply with open records requests, discovery motions, and subpoenas. This section further expressly recognizes that documents generated from the database are “summaries or reports.” This section should be used by lawyers when a TS claims that they cannot produce such summaries or reports—especially since they have been doing so for decades and have long claimed that these reports were the only information we really needed.
The reality is that these database summaries and reports contain information that does not appear on breath test slips and can form the basis for arguments that individual machines and tests have problems. DPS should not be allowed or permitted to dictate the form of court-ordered discovery contained in their database, especially when that information does not otherwise appear on breath test slips. Sections 4.4 and 4.5 go further and expressly recognize that database printouts do not always match the data on breath test slips, and that database reports can be generated to comply with open records requests and court orders. Section 4.6 contains a disclaimer that a TS is required to attach to database reports.
Part Five: “Breath Alcohol Instrument Calibration”
In common parlance, “calibration” is understood to be the process of programming a machine to accurately and precisely conduct its measurement function. Thus, for example, one could (at least partially) “calibrate” a bathroom scale by using the zero set knob to cause the weight designator needle to point directly at zero, as opposed to say the line for 10 pounds. While this would calibrate the machine to accurately reflect a result of zero when no weight was on the scale, true calibration of the scale would also require placing a known weight of say 100, 200, and 300 pounds on the scale and adjusting the scale so that it read 100, 200, or 300 pounds when that amount of weight was placed on it. In the scientific and metrology world, “calibration” encompasses not only the process of setting or adjusting the machine’s measurement function; it also includes a verification process: That is, once the machine has been programmed to “know” what amounts to 0, 100, 200, or 300 pounds, it is then necessary to determine how accurately and precisely the machine can measure other known weights within the zero to 300-pound range.15
In the Texas breath alcohol testing world, however, section 5.1 provides that “calibration” is used only to indicate the process of testing the accuracy, precision, uncertainty of measurement, and linear response of the machine’s measurements of known standards, and does not include the process of setting or adjusting how the machine determines its measurements. While this seems at least a bit peculiar, it is the product not only of misunderstanding, but also, and more directly, of a process built into the Intoxilyzer by its maker, CMI. CMI’s name for the process of programming or adjusting the Intoxilyzer to conduct its measurement function is called “Autocal.” Thus, the SOGs refer to an “Autocal” as the process by which the machine is taught what electrical response corresponds to various solutions, ranging from 0.00 to 0.40.
Section 5.2 provides that the “calibration” shall be performed prior to a machine being placed in service for the first time, any time the machine has undergone a calibration adjustment (an “Autocal” discussed below), or any other time as determined by the TS. Section 5.3 provides that a “calibration” shall be accompanied by the completion of the “Calibration Analyst Worksheet.” Section 5.3.1 provides that the “calibration” shall be accomplished with nominal reference values of 0.00, 0.04, 0.08, 0.16, and 0.40, in order (section 5.3.2), and that the solutions used for the process must be supplied by the Office of the Scientific Director. Section 5.3.1.1 provides that the Office of the Scientific Director shall make available in digital format the documents necessary to establish NIST traceability, unbiased estimate, nominal value, and combined uncertainty of the reference sample solutions.
Section 5.3.3 provides detailed instructions on how to perform the “calibration.” This procedure requires each simulator to be at 34.0 degrees centigrade 6.2 degrees centigrade and that the temperature be confirmed with a NIST traceable thermometer (5.3.3.3). The head space gas for each solution is tested 20 times (5.3.3.7), and each calibration check must be printed (5.3.3.6). The label from the nominal reference standard bottle must be affixed to the printed calibration check (5.3.3.7). The printed calibration checks must be signed or initialed and scanned into digital format (5.3.3.8). Only the final 15 calibration checks are recorded in the Calibration Analyst Worksheet (5.3.3.9). All results for the 0.00 solution must be 0.00 (5.3.3.9.a). The mean of the 15 analyses of the ethanol solutions must be within 0.003, or three percent, of the unbiased estimate of the standard (5.3.3.9.b). The standard deviation of the nominal 0.04. 0.08, and 0.16 solutions must be less than 0.001 (5.3.3.9.c). The standard deviation of the nominal 0.40 solution must be less than 0.002 (5.3.3.9.d).
Section 5.3.4 provides that if the process must be terminated before completion, all records will be maintained, and that a complete calibration procedure must be accomplished—i.e., they must start over, before a calibration Certificate is issued. Section 5.3.5 and 5.4 provide that once the calibration procedure is complete, all documents must be completed and signed and submitted for technical and administrative review by another technical supervisor using the Technical and Administrative Review Checklist. Once the process is complete and the calibration certificate is completed, all paper and electronic documents and files will be sent to and maintained by the Office of the Scientific Director (5.6).
Though the SOGs do not contain any specific requirements for the “Autocal” function, an understanding of it is necessary. The Intoxilyzer “Autocal” function is really quite simple. In order to “Autocal” the machine—i.e., adjust the prior calibrated values—the TS enters the password protected “Autocal” function on the Intoxilyzer. In preparation for the process, the TS will have prepared simulator jars with solutions ranging from 0.00 to 0.40 grams of alcohol per 210 liters of air at 34 degrees centigrade. There must be at least three different solutions and more is usually better. Four to five solutions are common. Once in the “Autocal” function, the machine will call for the TS to connect the jar containing only water, i.e., a 0.00 solution. The machine will sample the head space from this solution four times. The machine will exclude the first test result and will calculate an average of the electrical responses of the final three solutions’ measurements.
Basically, the machine has now been taught what electrical responses correspond to 0.00 grams of alcohol per 210 liters of air. This process is repeated with solutions that may include 0.02, 0.04, 0.05, 0.08, 0.10, 0.20, 0.30, and 0.40 grams of alcohol per 210 liters at 34 degrees centigrade, when prompted by the machine to connect that solution. Whatever quantity and values of solutions are used, the TS must use a 0.00 solution and a 0.40 solution. How many solutions between those values is left to the TS’s discretion, but two to three additional solutions are normal, and a 0.08 solution is most common simply because that is the threshold for per se intoxication. In each instance, the machine again samples the head space gas of each solution four times, excludes the first value, and calculates the average of the final three solutions. This “teaches” the machine what electrical response corresponds to the solution used.
The Intoxilyzers used in Texas also have a function that reports the results of the “Autocal” process. Almost all TS’s either turn off the function that causes this report to print or if the report is printed, destroy it. If the report is not printed, the machine destroys the underlying data and saves only a summary. Conceptually, one might reasonably question why a TS, with at least the tacit—if not express—approval of the DPS Scientific Director, would fail to print or would destroy the report produced by the machine. Though DPS posits other explanations, this author believes that these reports are not printed or are destroyed because the TS does not have the knowledge or expertise to explain these reports to a jury. In the 25-plus years I have been doing this, I have only had one TS actually turn over such reports. Copies of those reports are available on request. To make a very long story short, there is a great deal of data in these reports that appears to raise serious questions about the ability of this machine to consistently obtain similar results during the “Autocal” process: That is, the allegedly measured values often differ dramatically from the other results obtained by the machine. In any event, TS’s lack the knowledge and expertise to explain to a jury how—or even if—this report shows that the “Autocal” was successful or flawed.
Once the machine’s settings have been adjusted in accordance with the “Autocal” process, the value of an unknown solution introduced into the machine is determined by locating the point on the “line” where the unknown solution falls.
True calibration of the machine is not complete until the accuracy of the settings has been tested and confirmed. DPS and the SOGs refer to this verification process in part five of the SOGs as “calibration,” though, in reality, it is a check or confirmation of the calibration—i.e., the just-completed adjustment of the settings. This process involves again taking a variety of known and NIST traceable solutions and having the machine measure them. The goal is to see how accurately and precisely the machine will now measure a “known” solution. To test the accuracy and precision of the previously made adjustments (calibration or “Autocal”), the TS must again connect all the samples previously described, measure them repeatedly, and calculate the accuracy and precision of the results.
Some Ts’s use the same solutions to test the calibration as they used to calibrate or “Autocal” the machine. Doing so is fundamentally unsound, unscientific, and forensically indefensible. A simple example demonstrates the fallacy of using the same solution to calibrate and test the calibration. To teach (calibrate) my bathroom scale to accurately measure 250 pounds, I stand on it and adjust the measurement until it reads 250 pounds. Having done so, I then want to know how much I weigh, so I get back on the scale and see that it reads 250 pounds and conclude that I must therefore weigh 250 pounds. Of course, I have not proven that I weigh 250 pounds, just that the machine will repeat back what it was taught. If I want to know that the scale can truly measure 250 pounds, I have to use something different than what I used to calibrate it, and must know that the verification weight is truly 250 pounds. The same thing applies to checking the calibration on an Intoxilyzer.
Once the values from 0.00 to 0.40 are established, they are (conceptually) plotted on a graph in a way that establishes what is known as a linearity curve. Thinking graphically, a line is drawn through all of the plotted points, beginning at the 0.00 value and ending at the 0.40 value. Ideally, this should be a perfectly straight line. How close this line comes to actually being straight is called the “correlation coefficient.” Ideally, the correlation coefficient should be a value of 1.0, which would represent a perfectly straight line. As a practical matter, for various reasons, this “line” is never really completely straight (and is why it is called a curve not a line). To be clear, the “straightness” of the line should, however, result in a value that is very close to 1.0, with most scientific standards requiring a correlation coefficient of at least 0.999.
DPS does not, however, require any minimum correlation coefficient. One might speculate that no standard is set because it is doubtful that the Intoxilyzer could meet typical scientific standards under these circumstances. Indeed if the correlation coefficients that appear to be produced by the “Autocal” process are any indication, this machine does not have ability to consistently produce correlation coefficients of at least .999.
It goes without saying that all documents related to the “Autocal” and calibration verification process should be requested and obtained in discovery.
Part Six: “Thermometers Used in the Instrument Calibration Procedure”
Part Six establishes quality assurance guidelines for NIST traceable thermometers used in the calibration process of Part Five. In short, it is not enough that a thermometer (or any other device or solution) be NIST traceable at the time it was made. It is also necessary to ensure that the quality of the measurements remain valid over time.
Sections 6.1.1 and 6.1.2 provide that NIST traceable thermometers must be used in the calibration procedures and must be maintained in the TS’s calibration laboratory “which has limited access.” Section 6.1.3 provides that NIST traceable thermometers may only be calibrated by an approved vendor and once calibrated are suitable for use for one year. Section 6.1.3.1 requires that the vendor must be an ISO 17025 accredited laboratory capable of issuing a calibration certificate establishing traceability to a NIST reference standard.16
In essence, DPS has decreed that once a thermometer is calibrated, it needs no other attention or internal quality control testing for one year. This portion of the SOGs provides no internal method of measured quality control—that is, testing internally to ensure that at the time of use, the thermometer remains capable of accurately measuring the temperature. The SOGs also provide no requirement for the outside laboratory to report whether the thermometer was functioning properly when they received it at the end of the year. Without knowing that the thermometer was accurately and precisely measuring temperature both at the beginning and the end of the year, there can be no assurance that it was doing so at any time during the year.
Conclusion
DPS’s adoption of the SOGs is a good start. It begins to put “science” into a program that it has long touted as scientific but that has just as long lacked almost anything close to real science. But, it is just a start. DPS needs to beef up the SOGs standards, particularly where they are below industry standards, needs to stop the practice of destroying “Autocal” records, and needs to adopt similar standards for the many other aspects of the breath alcohol testing program. DPS also needs to change the culture that seeks to limit documentary paper trails. If they are practicing good science and following the rules, they have nothing to fear from a paper trail.
Defense lawyers need to obtain, read, and understand the SOGs. They also need to seek and obtain the documentary evidence necessary to confirm or rebut the proper operation of all aspects of the breath alcohol testing program. Judges need to understand that unless and until any laboratory, including a breath test lab, can document its claims of valid results and procedures, there is no validity. That documentation and verification are necessary components not only of good science but of the adversarial system is inherent in the documentation required by the SOGs. The SOGs require documentation and proof because there is no other way to prove valid results. That DPS seeks to do so in these new SOGs validates the decades of claims that they should have been doing so all along.
Ronald Reagan got it right when he said, “trust but verify.” This concept is as applicable to the legal process and forensic laboratories as it is to issues in everyday life in the rest of the world. Part of the inherent cost in running any type of forensic program is the cost required to verify to the outside world compliance with required standards. Anything less does a disservice to the pursuit of justice. All sides of the bench and bar should at least be able to agree that justice is only served with verified reliable science that we all can be confident in, and that the risk of injustice is substantial when any participant shirks the responsibilities to verify.
Notes
1. See Stevenson v. State, 895 S.W.2d 894 (Tex. Crim. App. 1995) (citing and discussing the pre-1994 version of Tex. Rev. Civ. Stat. Ann art. 6701l-5 § 3, which provided:
(b) Analysis of a specimen of the person’s breath, to be considered valid under the provisions of this section, must be performed according to rules of the Texas Department of Public Safety and by an individual possessing a valid certificate issued by the Texas Department of Public Safety for this purpose. The Texas Department of Public Safety is authorized to establish rules approving satisfactory techniques or methods, to ascertain the qualifications and competence of individuals to conduct such analysis, and to issue certificates certifying such fact. These certificates shall be subject to termination or revocation, for cause, at the discretion of the Texas Department of Public Safety.
This portion of article 6701l-5 § 3(b) was later codified in § 724.016 of the Transportation Code. Though the exact language of § 3(b) does not presently appear in § 724.016, all codifications of prior revised civil statutes were adopted without any substantive change, even when different or less language was used. Tex. Trans. Code § 1.001 (“The program [codification of statutes] contemplates a topic-by-topic revision of the state’s general and permanent statute law without substantive change”).
2. Prior to the adoption and release of the SOGs, though every technical supervisor I have asked under oath about the existence of a rules or policy manual applicable to breath testing—apart from what is contained in the administrative regulations—has testified that no such document or compilation of documents exists, one prosecutor has produced such a document that purports on its face to be “Forensic Breath Alcohol Laboratory: Office of the Scientific Director Policy Statements.” Resolution of what appears to be a conflict between the sworn testimony and the document I was given will have to occur in the cases still pending in which such testimony was given.
3. 37 TAC § 19.4(f). Approval of any breath alcohol testing program is contingent upon the applying agency or laboratory’s agreement to conform and abide by any directives, orders, or policies issued or to be issued by the scientific director regarding any aspect of the breath alcohol testing program.
4. See Harrell v. State, 725 S.W.2d 208 (Tex. Crim. App. 1986); May v. State, 784 S.W.2d 494 (Tex. App.—Dallas 1990, pet. ref’d).
5. All emphasis is added unless otherwise noted.
6. Harrell v. State, 725 S.W.2d 208 (Tex. Crim. App. 1986); Davis v. State, 949 S.W.2d 28 (Tex. App.—San Antonio 1997, no pet.). See also Atkinson v. State, 923 S.W.2d 21 (Tex. Crim. App. 1996).
7. 37 TAC § 19.4(g). Approval of a breath alcohol testing program may be denied or withdrawn by the scientific director if, based on information obtained by the scientific director, a designated representative of the scientific director, or a technical supervisor, the approved agency or laboratory fails to meet all criteria stated in this section.
8. In the words of the court, “Thus, assuming without deciding that the oral directive could qualify as a rule of the DPS in addition to the requirements of section 19.4(c), the terms of the directive as described by Finkley did not apply to Amend’s test.” It does not appear that Amend argued or that the court reached the issue of whether the oral directive constituted “any directives, orders, or policies” under 19.4(f).
9. “Autocal” is a function on the Intoxilyzer that sets and adjusts the calibration. The “Autocal” process is more thoroughly discussed below. On some occasions it is shown as “AutoCal.”
10. While it might seem strange to think that a TS would conduct business regarding an evidential machine at the TS’s home, some contract TS’s do just that.
11. The Texas standard of three percent is less demanding than the two percent standard required by the National Safety Council Committee on Alcohol and Other Drugs. The National Safety Council Committee on Alcohol and Other Drugs is an organization that sets industry standards for simulator solutions used in evidential breath testing. Interestingly, Mack Cowan, the Scientific Director of Texas’ breath alcohol testing program, was recently Chair of this committee. These standards may be found online at: http://www.nsc.org/get_involved/divisions/Documents/Handbook%20NSChistoryofCAOD.pdf. The two percent requirement is on page 99 of the document. Under this standard, a nominal 0.08 solution must measure to no less than .0784 and no more than .0816, meaningfully different than Texas’ lesser standard of no less than 0.077 and no more than 0.083.
12. ACA records are records reflecting an Air Blank, a Calibration Check, and then an Air Blank, hence ACA. The nomenclature “ACA” reflects the sequencing of the testing on those test slips, as opposed to the sequencing that appears on normal breath test slips.
13. Despite this new directive, DPS can and does still produce such reports when ordered by the court. This author received one on April 24, 2012.
14. Dee Wallace was a technical supervisor covering several small cities between Houston and Galveston. To make a long story short, over the course of several years, she routinely altered records to make it appear as though she had performed maintenance and changed solution when in fact she had not done so. Following discovery of her fraud, she was prosecuted and convicted. DPS invalidated several thousand breath test results as a result of her activity. More information about her can be found with a simple Google search.
15. It would never be scientifically proper to use the same items that were used to calibrate the machine to test the calibration. In the breath testing context, whatever solutions are used to “calibrate” the machine should not then be used to test or confirm the accuracy and precision of the machine’s ability to measure unknown values. Instead, other solutions of known (and confirmed) values should be used.
16. ISO 17025 is the global quality standard for testing and calibration laboratories issued by the International Organization for Standardization (ISO). This standard is what controls and forms the basis for ASCLD-LAB’s (American Society of Crime Lab Directors—Laboratory Accreditation Board) current “International” accreditation of forensic laboratories.
