A survey instrument was designed to elicit information from PHL directors or their designees regarding their interpretation and application of test results from CDC's MDDR. In addition, respondents were asked questions regarding how they first learned about the service, who in their jurisdiction is responsible for initiating requests for using the service, and customer satisfaction. CDC determined that this activity was public health program evaluation rather than research. Institutional review board approval for human subject research was not required. The survey was piloted by nine randomly selected PHL directors or their designees, who submitted samples to CDC's MDDR between September 2009 and February 2011. Feedback from this group was used to refine questions as needed and establish the estimated time required to complete the survey. The final instrument consisted of 18 multiple-choice questions and respondents were required to answer each question by selecting either one choice or all that applied as indicated in the survey. An open-ended response option was available for some questions. This data collection effort received expedited approval under an Office of Management and Budget (OMB) generic clearance package (Information Collections to Advance State, Tribal, Local and Territorial Governmental Agency System Performance, Capacity, and Program Delivery; OMB number 0920-0879).

The survey instrument was distributed electronically using Snap Surveys version Snap 10 Professional software [7] (http://www.snapsurveys.com) by emailing potential respondents a link to summit responses online. The sampling frame comprised 43 PHL who had submitted at least one isolate of MTBC to CDC's MDDR.

Most respondents indicated they first obtained information on CDC's MDDR at a professional meeting (26%), a CDC-assigned TB laboratory consultant (23%), or their jurisdictional TB program (20%). Only 17% of respondents indicated first becoming aware of the service through a formal communication (i.e., letter via email) from CDC. When asked who initiates test requests, PHL officials collectively indicated that most requests originated from the jurisdictional TB program (71%) followed by the laboratory (49%) and health care providers (34%). Over 97% of PHL officials indicated they were either very satisfied or satisfied with the turnaround time for receiving test results.

Over 74% of PHL officials indicated that molecular results from CDC's MDDR are provided to health care providers in less than two business days after receipt of the report. However, 23% of PHL officials indicated results are reported to health care providers by their TB program and not the PHL. In these instances, the time frame for health care providers to receive molecular results could not be determined using this survey. Ideally, with expanded access to electronic reporting, molecular results could be provided simultaneously to submitting laboratories and for population of electronic medical records to avoid potential delays in initiation of effective treatment regimens.

PHL providing results from CDC directly to health care providers did not withhold the interim report of molecular results until phenotypic DST was completed by either CDC or their own laboratory. When reporting results to health care providers, PHL most often provided a copy of the CDC report (89%) but frequently verbally communicated results as well (26%).

Thirty (86%) of the respondents indicated that they always compare molecular results from CDC's MDDR with phenotypic DST performed in their own laboratory with the primary intent to identify any discordant results for first-line drugs. Of the 32 PHL who compared molecular results from CDC's MDDR with their own phenotypic DST, 16 (50%) reported to have observed discordant results. Most PHL observing potential discordance took multiple actions with the most frequent being to notify their TB program authorities of the discordant results (81%), retesting the isolate in their own laboratory (63%), and contacting CDC directly to discuss the results (56%). Two PHL reported taking no additional action when they observed potential discordance. Of 15 PHL that performed second-line phenotypic DST in-house, 13 (87%) always compared their results with molecular results from CDC's MDDR. The most frequent action selected among those performing second-line testing was to notify their TB program of the potential discordance (83%). Twelve of 35 PHL (34%) responding observed discordance between the molecular results and phenotypic DST performed and reported by CDC's MDDR.

The high frequency of observed discordance contradicts findings from a recently published study where molecular results from CDC's MDDR were compared to local phenotypic DST results collected from submitting PHL [6]. This previous report found 90.1% concordance between CDC molecular and local phenotypic DST results. Discordance between molecular testing and phenotypic DST was due to not detecting mutations in loci associated with resistance in isolates that were later determined to be drug resistant by phenotypic DST. However, this prior study only compared molecular and phenotypic results for RMP and INH. In the present study, respondents needed to consider discordance between CDC's MDDR molecular results and their local phenotypic DST results for detection of drug resistance for all first-line and second line drugs used in testing procedures. This would increase the odds of discovering discordance between testing methods. When potential discordance was noticed by PHL, nearly all contacted their TB program but on occasion some took no further action. This circumstance requires further inquiry because discordant laboratory results should be addressed in the process of clinical decision making.

PHL officials were asked if there was any impact on their own local phenotypic DST when the first available results were the molecular results from CDC's MDDR. Of these, 24 (69%) indicated that there was no impact on their own phenotypic DST. Three PHL (9%) acknowledged that when resistance was indicated by molecular results, they would refer the isolate to another laboratory other than CDC for additional testing. Among the open responses to this question, three PHL officials (9%) indicated they would initiate additional second-line phenotypic DST if the molecular results from CDC's MDDR indicated drug resistance.

Since CDC's MDDR uses agar proportion for phenotypic DST that can take five weeks or more to complete, PHL may choose to use a more rapid phenotypic method to confirm drug resistance and not wait for a final report from CDC. In addition, PHL may be seeking information about additional drugs not in the panel used by CDC's MDDR at the request of jurisdictional TB programs or healthcare providers.

Twenty-six (74%) respondents reported molecular results were either not difficult or very easy to interpret. PHL officials were also comfortable discussing results with either healthcare providers or TB program staff. Among 28 PHL officials contacted to help interpret molecular results, 23 (82%) were comfortable in these discussions. When asked what resources they sought for help with interpreting results, 14 (40%) responded that they did not seek help. For 21 PHL officials who did seek assistance with interpreting molecular results, 16 (76%) contacted CDC. PHL officials sought help less frequently from other sources. Seven (33%) reported doing their own research to find information on molecular testing.