Maximum Advantage Reagent Traceability System (MARTS) is a high-performance database for tracking purchased reagents and solutions prepared in your laboratory. MARTS maintains a list of all the chemicals in your facility, tracks location and quantity of reagents, monitors expiration dates, generates reports listing chemicals close to expiration date, and quickly accesses Material Safety Data Sheets (MSDS). MARTS eliminates the need for numerous paper logbooks and keeps information accessible and accurate.
Maximum Advantage Basic Quality Control Software (MA Basic QCS) was developed for smaller laboratories that analyze a limited number of parameters. MA Basic QCS provides a manageable and comprehensive quality control data-entry form. MA Basic QCS confirms that your blanks, spikes, duplicates, laboratory control standards, and continuous calibration verifications are within control limits.
"Products and services from Alloway have provided us complete confidence in responding to any audits. As a result of Alloway's assistance, we now have a manageable quality system. Also, their software has been exceptional for tracking and graphing QC data. Alloway has been a great partner with very affordable prices. They really care about our lab's success."David BeachFindlay (Ohio) Water Pollution Control Center
Alloway is a full-service environmental laboratory specializing in analyses on drinking water, wastewater, and solid and hazardous waste. In addition, Alloway offers innovative software, laboratory development, and training options. With three locations to serve you and more than 30 years of experience, Alloway is your resource for defensible data.
We contracted the services of a Canadian-based software developer with experience in the development of online surveys and an understanding of the health sector ( ). Our requirements were complex; providing respondent and interviewer access to the survey using a variety of computer systems located within diverse environments across geographically distributed locations and all systems requiring capacity to securely upload individually gathered data to the master dataset located on a remote server. This is typically where Internet accessible surveys would be well suited; however Internet access in our settings was inconsistent and in some communities nonexistent. Without reliable Internet access, our best solution was to purchase several laptop computers (for use by interviewers undertaking data collection concurrently in each setting) and have the survey software installed on each laptop. This allowed for offline survey completion and temporary storage, with subsequent upload of data using a secure file transfer service when internet connectivity was available. Each interviewer was allocated a unique identifier, and each data file had a unique file naming convention.
Another important consideration in the software development process was the data upload procedure. Development, refinement, and testing of this process took into account the steps required to connect to the server, how soon after data collection the data had to be uploaded, and the type of confirmation that was generated to signal a successful upload. Additionally, system checks were put in place to ensure the same data could be uploaded only once. Testing of the survey and the upload process to ensure each was fully operational and behaving as expected involved internal and external review phases. The internal phase involved a cycle of development, review and testing, and modification, followed by further review and testing. This process took into account the survey content, visual appeal of the survey, and ease of navigation. The former involved checking the question order, completeness of questions, spelling, grammar, and punctuation. The latter involved consideration of overall appearance of the survey, the colors, the format and layout of responses, the number of questions per page, how the questions were separated (different colors and width of lines), the ability to advance or move back in the survey, and the ability to change responses. The external phase involved testing of the software by healthcare aides and evaluation of the overall appearance, functionality, and ease of navigation within the survey.
The data teams participated in intensive interviewer training to ensure standardized technique and the collection of high-quality data. To facilitate this process, an interviewer (procedure) manual and an interviewer quality control protocol were developed and implemented as components of the data quality control program. The interviewer manual contained technical information on the TREC study, the survey, the step-by-step process of conducting a CAPI interview, and an overview of the CAPI software and the processes by which the data were to be handled. The interviewer quality control protocol (Supplementary File 1; see supplementary materials available online at ) was central to the quality control program and contained three core components: (1) characteristics of a successful interviewer, (2) training, and (3) tracking and monitoring processes.
(1) Characteristics of a Successful InterviewerFour broad categories of characteristics of a successful interviewer were identified based on a review of existing literature and our experience with conducting face-to-face structured interviews. The four categories were (1) physical attributes, (2) personal characteristics, (3) technical skills, and (4) compliance with interview procedures. Physical attributes included open posture, consistent eye contact (with interviewee), and comfort with conducting the interview. Personal characteristics included a personable demeanor, engaging with the interviewee, appropriate speed of talking and clear and audible speech, appropriate (professional) dress and hygiene, and ability to problem-solve (e.g., technological problems) during interviews. Technical skills included ability to log on to the computer, ability to open and launch the virtual server CAPI software, ability to navigate through the survey, acceptable typing speed, ability to conduct the interview while entering responses with minimal delays, and ability to connect to the virtual server to allow data synchronization and upload following the interview.
(3) Monitoring and FeedbackInformation on the quality of the survey data collected in the CAPI interviews and the process of conducting the interviews was monitored throughout the data collection period. Monitored information included survey findings (e.g., missing data, skewness) and process-related data (e.g., travel time, time on site, number of interviews completed/in progress/refused) collected using standardized forms, which were submitted to and verified by the central office for the TREC study. In the event of discrepancies or errors with the process data, the data manager for the study would contact the research manager for the indicated province for resolution. Once verified, the information was entered (and double checked for accuracy) into a statistical database where it was analyzed and used to generate quality reports. Security and confidentiality policies were enforced for all reports (e.g., forms had to be sent by bonded courier; courier packages had to be received by an identified person in central office and were documented and stored in a locked cabinet). Also as a part of the quality control program, following each interview (once the respondent (healthcare aide) left the room), interviewers were asked to complete a series of questions (the interviewer checklist, Supplementary File 1) on the interview process. This also allowed for a better understanding of the circumstances in which each survey was completed. These data were analyzed regularly (quarterly) to further assess quality of the interviews and compliance with the quality control interviewer protocol. This information was fed back to the interviewers when necessary.
The third and final phase of our quality control program was a data cleaning and processing protocol (Supplementary File 2). The protocol consisted of six steps and was implemented quarterly by a data analyst for the study: (1) systematic data entry, (2) data cleaning, (3) prederivation processing, (4) derivation of scale scores, (5) descriptive assessment of derived scores, and (6) assessment of missing data. Throughout this process, a four-part report was produced for the study lead investigators: steps 1-2 (report A), step 3 (report B), steps 4-5 (report C), and step 6 (report D). Each report was reviewed and approved by the study principal investigator before the data analyst proceeded to the next phase of cleaning and processing.
Process-related data collected from the interviewers included, for example, whether or not a paper-based survey was used. These data indicated that the proportion of paper-based surveys used was small and fell over time (from 2.9% in Year 1 to 1.8% in Year 2; chi-square, ): an indication that the procedure for enabling software updates overnight (to prevent computer start-up delays during the daytime) was functioning well and the interviewers were confident in using the software. Data were also collected on whether or not the interviewers were able to set up the interview according to protocol. Overall, in only 3.4% (Year 1) and 4% (Year 2) of cases was the interviewer unable to set up in accordance with the protocol. On most occasions, the interview was conducted in a private location as per protocol (72% in Year 1 and 78% in Year 2). Frequently the location was also visible to other staff (65% in Year 1 and 75% in Year 2) and close to resident care (68% in Year 1 and 86% in Year 2). Interruptions during data collection, which could potentially threaten the quality of the data, were also monitored. The majority of the interviews were conducted without interruption (76% in Year 1 and 84% in Year 2). Another possible threat to data quality was pauses (where the interview had to be stopped and restarted). The majority of interviews proceeded that required a pause (91% in Year 1 and 95% in Year 2). Interviewers were also asked to rate the overall quality of the interview from 1 (terrible) to 5 (wonderful). Their perceptions of overall quality improved between Year 1 (mean 3.84) 2 (mean 4.11); this improvement was statistically significant (chi square, ) which could reflect improved competence and confidence from training and feedback provided in the data quality control program and as they gained experience in conducting the interviews. 2b1af7f3a8