Tobacco Documents Online

NTO0123125 MICROBIOLOGICAL ASSOCIATES JUSTIFICATION FOR THE PURCHASE OF A HP-1000 SERIES E MINI-COMPUTER FOR THE CTR/MA CIGARETTE SM OKE EXPOSURE PROGRAM Prepared by: Carol J. Henry, Ph.D. Richard E. Kcuri, Ph.D. Miles Haven April 1, 1979 ~James M. Ludwick Director of Contract Administration Richard E. Kouri, Ph.D. Director of Research

I HT®0123126 TABLE OF CONTENTS Page I. Present System 1 II. Problems 2 (II. P,oposed Sotutions 3 IV. Conclusion 8 V. Time Schedule 10 VI. Budget 11 VII. Appendices 12

H.T90123127 I. Present System In 1973, the "Experiment Information Management System" for CTR-related studies at MA was designed and implemented. The system was desianed to manage long term In vivo laboratory animal experiments. In general this system work-; a s follows: a) A nimals are randomized, given specific identi- fication ear t ags, and assigned to a specific experiment, group, and treatment. b) Forms are filled out by the technician assigned to this study and these forms are keypunched, verified, interpreted by our edit program, and added to the data base via an update program. c) As the experiment proceeds, inoculations, observations, weighings, bleedings, tissue sampling and biochemical analyses are made and added to the data base as described in Section b). d) At the time of death or sacrifice, autopsy forms are Implemented to describe any gross pathological observations and to assign specific histopathology numbers to oach animal. These forms are in quadruplicate so that one resides with the central file, one goes to the histology department and two go to the pathology department (one is returned). e) The computer system supports SNOP (Systematized Nomenclature of Pathology) codes for the interpretation of the pathological results. f) This system can support up to 99 separate on- going experiments, which could include information on a total of 120,000 animals or over 30 miilion characters of information. 1.

Ht80123128 Data are constantly updated so that retrieval of all pertinent information is possible at all times. During and after the course of the study such reports as tumor incidence, tumor free survival, autopsy, geneology, and statistical analysis of these reports are genErated by the computer. At termination c•f a study, these data are not,r placed on an archive file where at any later time, they may be brought back on line and evaluated or compared to other studies. 11. Problems Data integrity on our present system is assured by the system design, edit, update, and report programs. No animal can get into a final report without going through at least four different levels of verification. This system works well when the process of data collection iF fairly straightforward and the amount of data collected is such that it can be handled manually and entered into the system by conventional methods. The data generated in our lst smoke exposure study showed clearly that the amount of data collecte d should be Increased to include such factors as a) smoke generation data, b) carboxyhemoglobin analysis, c) an infinite number of addltional remarks (presently we are limited to six), and t:) additional gross and histopathologic observations. These additions are currently being made and will increase the amount of data collected by a facto r of fou r. (An outline of the proposed final integrate d file is given in Appendix A). The smoke generation data are very important not only for immediate interpretation for usage with the safety systems (to be described in the next paragraph), ~ 2

I HT®012312: but also because some permanent record of these data should be made so as to verify that the exposure conditions stipulated by the protocol are in fact being carried out. Thus, although a puff-by-puff profile of total particulate matter (TPM) after every smoke exposure treatment is now being stored on strip charts, there is no presently avai 1- able system to verify and authenticate that the correct smoke regimen has been given to the correct set of animals. A second observation made during this first long term smoke exposure study was "that some method of on-1 ine monitoring of smoke gene ration data was required for our safety systems. As mentioned previously, we are presently monitoring the TPM levels on strip-chart re corders and thus have a puff-by-puff profile for every exposure carried out at MA. What seems to be necessary is an immediate feedback to our safety systems so that such factors as air or smoke flow rate and TPM can be instantly monitored, evaluated and decisions made as to the relative risk this specific exposure regimen may present to the animals. This on-line decision capability should decrease the "Smoke Exposure Machine" (SEM)- associated death rates measurably. These aforementioned factors demand a methodology whereby the informatibn and d&ta generated during these smoke exposure studies can be collected, evaluated, pro- cessed, and stored automatically at the time these data are generated. This requires the use of what is termed "real-' time" applications and thus the integraticn of d "real- time" computer with the SEM would be the most dirPct and efficient way to support the current experiments. Our present computer (a Hewlett Packard 3000 Series III) is a "time-shared, general purpose" computer rather than 3.

HT60123130 "real-time" computer and thus is not capable of interpreting analog signals and acting upon the collected information at the instant it is generated. A real time compute r has this capibility. III. Proposed Solution We propose that a Hewlett Packard 1000 Serie s E computer be purchased and integrate d with the 2 SEMII's and the 8 smoke/ai r channels on the two animal contai nment units. This configuration should allow for the following: A. Smoke Exposure Documentation. 1. Date.' 2. Checkout Procedure. (a) Puff volume, pressure normal. (b) Flow sensors functioning and all hoses connected. (c) Electromagnets functioning. ~(d) Safety alarms funct i oni ng . 3. Experiment Number. 4. Group number and/or mouse numbers for each strain of mouse and on which channel. 5. C i ga re t te type, numbe r of c i ga re t te s scheduled and numbe r of puffs per cigarette. 6. Exposure time (program number). 7. Smoke concentration (dilution and purcie settings) ISSUE WARNING IF DATA ABOVE (c-f) DOES NOT CORRELATE WITH INFORMATION ON FiLE FOR EXPERIMENT). 4.

I H T®01 231 3'~ 8. Time, temperature, and humidity. 9. Channels used (A + D, or B + C, etc, ). 10. Number of mice per channel. 11. Run numbe r. 12. Any changes being made in exposure for this day and documentation of the reasons for change. 13. Minimum flow rate at end of each channel. 14. TPM generated for each channel and e ach exposu re . 15. Mouse number of any mice which died during an exposure and the description of the circumstances associated wih the de ath. 16. For experiments wi th sequential, mul tiple exposu re, docume ntat i on of each exposure and a summation of total TPM per day for each group or individual mouse. 17. Be able to retrieve exposure history for each group or individual mouse. 18. Integrate exposure history into the d ata base for the treatment, observation and weight records, blood or urine analysis, and pathologic diagnosis. B. Safety systems to be controlled by the "mini- computer" and when activated would issue a wa rn•ing, then shut down the SEM and open the animal modules. 5.

1 H T®0123132 1. aocument malfunctions with the SEM: ejector, lighter, distribution valve, puffer valve, hose connections, blower, loss of power, dome pressure (puff by puff), 2. Monitor safe flow rates to the mice through the thermistors - document any opening of the modules.. 3. Monitor puff by puff TPM record for any abnormalities (shape, duration, etc.). 4. Monitor how far cigarettes are burrriing (<23 mm). 5. Monitor the safe levels for the TPM/puff or the TPM/exposure and activate the modules when necessary. 6. Monitor flow rate before and after module, C. Protocol reevaluation will be executed daily to interrogate all scheduled activ;ties. Such parameters as the following are proposed to be monitored. 1. 'fhe genetic and physical characteristics of the test animal. 2. A randoming procedure for animal allocation to experiment. 3. Environmental conditions in the holding rooms. 4. Microhiological surveillance requirements and response procedures. 5. Food, water, standards and requirements. 6.

H T®01 23133 6. Routine and special handl ing instructior.s . 7. Experimental variables to be recorded. 8. instructions for special handling or re- moval of animals. 9. Instructions for consolidating cages and/or racks i n expe r i ment rooms. 10. Schedul ing of special observations or actions. 11. Genetic trace and census requirements. 12. Routine and special pathology procedures re qu i re d. 13. Validated statistical methods for treatment of data. 14. Data access approval, audit, and correction procedures . 15. Anticipated report formats; and 16. Schedule of anticipated scientific decision poi nts . It is proposed that the purchase, installation, and monitoring of the SEMII's by the HP-1000 mini-computer be completed by October 1979. Pert i nent data wi 1 l be col lected and stored on the HP-3000 unti l that time. Data col lected and acted upon by the HP-1000 wili be transferred to the HP-3000 (where the complete data base Is stored) every night via a tape cassette. By December 1979, software for a communication link will be available from Hewlett Packard so that the two computers can be directly connected over a telephone line. Presently avai lable equipment from HP is 1 imi ted to 2000 ft. for thi s di rect connection and so cannot be used for this proposed configuration. 7~

Ht®4123134 IV. Conclusion The breadth and depth of ongoing experiments at Microbiological Associates, the instantaneous monitoring of the SEM safety systems, and the demands for more sensi- tive analytical techniques, verification and documenta tion of our studies (see Appendix B for amount of data gene rated on CTR-101), all have contributed to increase the data burden on our laboratory personnel and computer hardware configu- ration. We wish to begin a transformation from handwritten, error prone means of data capture to a sophisticated, labor saving and more accurate automated data collection system. Such a system will promote confidentiality, allow accou ntable audit trails that the protocol was followed, and provide cost effective methods of data collection, processing, and reporting. MA's data processing department will continue to: ® be responsive to CTR user requirements in- cluding the collection of data and the gen- eration of timely reports in desirable formats. ® provide maximum verifications and editing of ;nput data. 0 provide authorized personnel with an audita ble means for error correction. ~ allow tracing of an animal or treatment group throughout the entire course of the experiment. . maintain data collection storage, and reporting procedures that are well documented and con- sistent with the experimental protocol. 8.