SUMMARY OF DISCUSSION - AVSS TAG MEETING
Sacramento, California
May 16, 2000

1. Introduction By Attendees and Opening Statements.

1. Review of minutes from November 8, 1999 TAG meeting.

2. Version 4.9 Updates.

Version 4.9 was released on November 9, 1999 to be ready for the new birth certificate forms and to update all 250+ AVSS servers throughout the state before January 1, 2000. The challenge was met according to schedule.

1. The most important change in Version 4.9 was related to AMR 99-026, which allows up to three race responses for mother and father. Race choices were restricted to a CBC list containing nearly 900 values. There was little negative feedback from LRDs or hospitals. There were, however, some problems with missing values for multiple races among imports from foreign systems (Kaiser and Site of Care).
2. The new birth certificate forms (VS-10D, Rev 1/00) proved to have more consistent vertical spacing than previous ones (Rev 5/97). The AVSS Project also made changes to the HP LASERJET FORM ALIGNMENT suboption to permit vertical adjustments in 1/10 inch increments. As a result, there were no problems in printing the new race values on the new forms. There was, however, a problem related to the new ability to print Field 17 (Date of Registration) under the LFN ASSIGNMENT suboption. This was due to several reasons: variations in top margins on the new forms, variations in laser printers, the minimal space available to print the LFN barcode on the lower margin, the lack of user-adjustment for vertical positioning of the LFN, and non-optimal spacing between the LFN, its barcode, and Field 17. Alan Oppenheim reported that LRDs were informed that it is permissible for hospitals to retain the privacy notification (bottom portion) of the birth certificate, so as to allow more space for printing the LFN barcode. AVSS was modified to optimize the spacing of the printed fields and to allow user adjustment of the initial vertical offset. These measures appear to have solved the problem.
3. There was a problem with miscoded races for AVSS/CMR due to the change reflected in the CMR card going from one race field that included Hispanic to two fields: one for race without Hispanic and one for ethnicity, composed of Hispanic subchoices. There was a miscoding error when the latter two fields were combined to create a single field that would maintain consistency with the historical single race field variable. There was also an inadvertent change in Version 4.9 that made Field 36 (Mother's Middle Name) on the Death Record a required field, whereas it should be optional. Also there as a need to add 'SCAT' to the list of choices for Field 41 (Type of Disposition) on the Death record. Because of the need to correct these problems, an updated Version 4.9A was prepared. Since the problems did not affect hospitals, only the LRDs were updated to 4.9A.
4. Since that time there was one significant problem uncovered that will necessitate a hospital update: if a ^F or ^S command is used during the multiple race prompting, the record will not be marked as incomplete. Thus, when the user attempts to complete the remaining fields AVSS will not automatically prompt for them so the user must edit each subsequent field, and could overlook one or more. Although this occurs infrequently, it will necessitate another update to all hospitals and LRDs. The new version will be called 4.9B. Several other less significant corrections will be included in this update.

3. AVSS Modification Requests (AMRs): Review and Discussion.

99-120: OK to add OTHER to list of Salmonella Serotype choices in the CMR. Also allow the user to specify; e.g. OTHER:SPECIFY.

99-128: OK to add the choice of ~DELETE to the DISEASE list in the CMR form. The tilde (~) prefix will force this choice to the bottom of the list in ASCII sorting sequences. CMR reports will need to be modified to discount such cases. The original ICD code will be retained so that the original disease choice will be preserved. It was decided that a new CMR menu option called (for example) MARK RECORD FOR DELETION would reset the value to ~DELETE automatically. Gail Cayton suggested that marking such cases at the state level should result in the records being automatically reallocated to the LRD. After some discussion this concept was approved. It was agreed that Sacramento County should be the beta test site.

99-130: LRDs with dialup hospitals and a telephone 'hunt' system would like to have user name displayed as part of the SYSTEM STATUS display. While this is not possible using the %SS routine, which is a MSM-written utility, it is feasible for AVSS Project programmers to create an alternative. If that is possible, the entire %SS display should be optimized to meet LRD needs.

99-136: It was suggested that the need to accommodate one-letter codes in order to report 'Any Mention Of' could be accommodated by using 107A (Cause of Death). There was no consensus on adding another prompted field to the CDC record. After some discussion it was agreed that a new field called AMO would be added to the CDC, but not prompted. This would allow LRDs who wish to add this field to do so by using the EDIT CERTIFICATE suboption while not imposing an extra prompt on other LRDs.

00-019: The 'radical' solution of requiring all hospitals to use the LIST TO BE SENT TO HEALTH DEPARTMENT was approved by all those present, understanding that records from foreign systems would be exempt. This might be included in Version 4.9B. Riverside County volunteered to be a beta test site. Before it is implemented, LRDs should poll all their hospitals to be sure they are using the LIST. This AMR would improve the quality of data, conform to the historical pattern of incrementally tightening quality controls, and in the worst case, the LRD would have to occasionally run the LIST.

00-015: OK to develop a foreign import capability for CMR as described in this AMR.

99-069 (revisited): Michael Rodrian and Alan Oppenheim reported that it is OK to allow AVSS to print the health officer's name in Field 16 (Local Registrar's Signature) as part of the LFN Assignment process. Ventura County will serve as the beta test site.

4. The Future Of AVSS.

1. Ron Williams briefly described the history of AVSS, beginning in 1980 and continually funded since then by federal and state MCH grants and contracts. What started as primarily a research project with some computer development has now become primarily a computer project with much less research activity. About one-half of the funding now comes from MCH with the remainder from LRD technical assistance fees. The current 3-year state MCH contract will expire on June 30, 2002, at which time Ron plans to retire. He met with Peter Abbott, Alan Oppenheim, and Michael Rodrian on February 15, 2000 to discuss the future of AVSS after his departure. It was agreed that DHS would attempt to hire AVSS technical staff as state employees. Michael Rodrian reported that MCH will likely redirect funding to OVR, but a budget change proposal must be written and approved, and he is committed to that process. Whether and how LRDs would continue to pay technical assistance fees (to DHS) is unclear at this time.
   1. Concern was expressed about the future role of TAG. Michael Rodrian stated that TAG would continue, especially since the LRDs will still be the beginning point of vital records registration. He is committed to maintaining the high standards for the current system and sharing data as well as decision-making with the LRDs.
   2. Deborah Martin suggested that LRDs be reimbursed at $1.00 per record for the data being transmitted to OVR since the state passes it on to NCHS and collects associated fees. In principle, this federal money should be shared with the providers of the data. Ron Williams stated that AVSS could not have survived without LRD assistance and they indeed deserve considerable credit for their past and ongoing contributions. Michael Rodrian responded that while it was unlikely that funds could be returned to the LRDs, there might be a way of returning in-kind services.
   3. There was a question from Kyle Lumen about retaining the two principal AVSS programmers (Peter Chen and John Marinko). Ron Williams acknowledged their unsurpassed loyalty and devotion to the AVSS Project and felt that they would likely give adequate notice so as to train replacements should they decide to leave or retire. With M classes being taught at UC Davis, it would be more likely to find replacements in the Sacramento area than in Santa Barbara.
   4. Ron Williams presented his view of the future of the AVSS system architecture. AVSS began with users as terminal-based clients in the early 1980s. By 1986 the first standalone PCs were being installed in hospitals and this trend accelerated, especially in Southern California. In his opinion the distributed PC model has been more problematical and difficult to maintain than the time-shared approach. He demonstrated an inexpensive 'AVSS Appliance' having a small footprint that could replace PCs. Ron believes that the future of AVSS lies with the client-server model, with thin clients at hospitals and a central server at a single location.
   5. Pam Isaac asked when AVSS would be available in the Windows NT environment. Ron Williams reported that it is the AVSS Project policy to exhaustively test a new platform before releasing it. AVSS/NT was installed at OVR in November 1999 and has been performing very well. However, it is unlikely that a LRD version will be released without first considering the Internet approach since technical support for the AVSS/NT will be considerably more complex and expensive. Since it is likely that DHS will be providing technical assistance for AVSS after 2002, it is appropriate that it be involved in the decision on where, when, and how to implement AVSS/NT.
   6. Pam Isaac expressed concern as to whether the central server model would allow LRDs to access their data as is currently the case. Ron Williams responded that the Internet model should look like the current distributed model, and, if necessary, data could be sent to the LRDs on a periodic basis. Pam responded that the state often charges fees for such services. Michael Rodrian felt that was presently the case since there are processing and data analysis costs required to produce other datasets, but in the case of returning AVSS data, there would be no costs, and thus no fees.
   7. A question was raised about the future of the AVSS software license. Ron Williams reported that that issue is being studied by University attorneys, including the Office of Technology Transfer, with three possibilities: the Regents retain the license and continue to charge fees according to the existing formula (possibly revised to account for the Internet), the license is sold to DHS, or the software is made public domain. The latter is unlikely.
   8. Joe Quintanilla asked about converting from VAX-based AVSS to NT since Orange County's VAX is becoming less reliable. It was noted that Fresno County is upgrading their VAX hardware and M software (Cache'), and that would suggest that it may be possible for VAX systems to continue operations until the Internet version is explored.

5. Vital Records Revision.

1. Jane McKendry reported on the vital records (birth, death, and fetal death) revision process, which occurs about every ten years. The process begins with NCHS setting a national standard, but each state may vary in the extent to which they meet that standard. Jane has been gathering data on what various parties would like to add or delete from the current certificates. Data elements must be finalized in about one year in order to prepare AVSS and other systems. Work groups will be established and meetings scheduled. Changing data elements on the birth and fetal death certificates will require legislation. The current NCHS model birth certificate spans two pages, but it is likely the California will limit its certificate to one page. It is possible that California could gather additional data items electronically but not by paper, like the current MAR variable for example. Those individuals wishing to participate should contact Jane McKendry at cmckendr@dhs.ca.gov or 916-445-6355.

6. Electronic Death Registration System.

1. Natlee Hapemen reported that the AVSS/EDR pilot is continuing in Santa Barbara County with about a 70% participation rate for the two active funeral homes and about 35% overall. One other funeral home is using it sporadically and another has requested installation.
2. Alan Oppenheim reported that NAPHSIS has received funding to review EDR systems throughout the US, and SSA staff visited Santa Barbara on February 14-16 to review progress on AVSS/EDR. The AVSS Project and Santa Barbara County gave presentations. A copy of the presentation overheads was contained in the TAG handout. A report from NAPHSIS will be forthcoming soon. It is expected that an Internet model will be recommended.
3. The CFDA is developing and testing an approach to electronic certification, called PCL (Physician Certifier Line), that will use biometric voice printing. DHS will fund up to $10K through Santa Barbara County to the AVSS project to see if an interface between AVSS and PCL can be approached. Peter Chen will be in charge of that project. If this is successful then it would be necessary to develop a larger project to integrate PCL into AVSS.
4. Ron Williams reported that he has contacted UC attorneys who have stated that there would have to an indemnification agreement between DHS and UC attorneys agreeing that UC would not be liable for damages in case of the inappropriate/erroneous/fraudulent use of PCL to certify death certificates printed by means of AVSS.

7. Confidential Morbidity Report (Discussion Led By Mark Starr: See attached CMR Topics).

1. AMR 00-015 as discussed above would serve to establish an AVSS standard in every LRD regardless of the type of system that is eventually used for primary data entry of CMR cards. This will facilitate the efficient transfer of CMR data to the DHS/DCDC system and produce uniform reports throughout the state. The variable record length format will allow foreign systems to pass whatever information they have available for all cases, assuming a minimum number of required data elements are available.
2. New staff at DCDC: Scott Schmolke (replacing John Hastings) and Rosalie Trevejo.
3. Mark Starr can be contacted at mstarr1@dhs.ca.gov

8. EAB/NANA.

1. Michael Rodrian reported that SSA was going to impose an administrative order to issue a SSN for the child only if at least one parent's SSN was reported. However, a recent (5/3/00) memo from SSA appears to negate that policy. There is considerable confusion on the matter at present. Rod Palmieri stated that if the parents' SSNs were missing or invalid a child's SSN would be issued, but there would be followup.

9. Bar Coding of LFN and SFN.

1. Orange County recently began bar coding but had printer problems and a new one is on order. A releatively inexpensive high performance laser scanner (PSC Quickscan 1000) is now available for slightly over $200 as posted on the AVSS web site under Hardware and also listed in the TAG handout.

10. AVSS Automated Birth-Death Matching.

1. MCH is providing funding to improve AVSS birth-infant death matching and to possibly incorporate it into the ongoing OVR operations. Matching is preferably performed at the state level since family mobility can result in a change in residence during the first year of life and consequently infant death certificates will not be reallocated back to the LRD of birth. When AVSS attempts to match such an infant death at the local level, the birth certificate will not be in the LRD database and thus cannot be matched. This limitation does not exist at the state level, where a 95% match rate for 1999 infant deaths was attained. The remaining non-matches were frequently cases that need further investigation due to unusual circumstances surrounding the death. There were also cases where the age of death was in error on the AVSS death certificate since the CDC record is batch entered by the LRDs. A new 'cross-match' index (BD) was recently created at OVR. This index provides a linkage between AVSS birth and death certificates, regardless of age. Following the completion of birth-death matching at the state, this index and all infant CDCs could be distributed to all LRDs by means of a CD.
2. Deborah Martin asked about the possibility of producing a birth cohort file using AVSS. Ron Williams responded that since most of the CDC mortality data was incomplete and unverified, he believes that there is little useful infant mortality information other than the age of death. Since this information already exists in position 813 (Death Indicator) on the CBC 1400 byte record, that format should suffice to produce neonatal and postneonatal mortality statistics. The problem is how to set this field when there is no date of death reported in Field 15A for those births that AVSS has matched to deaths. There will be a need for OVR to establish a policy on the proper procedures and how to return the appropriate death indicator information to LRDs.
3. Several participants suggested that it would also be useful automate fetal deaths. The AVSS Project will investigate this possibility.

11. Equipment/Operating System Recommendations.

1. AVSS is a DOS-based program and it is ideally hosted on a low-cost dedicated computer running DOS 6. If it is necessary to run on Win 9x, then the procedure posted on the AVSS web page must be followed.
2. As reported above, there is now an NT version of AVSS. This network version is expensive to install and maintain, and is inappropriate for hospitals. The AVSS Project is also concerned about installing it at LRDs and will defer to OVR in this decision since UCSB will not be supporting LRDs after 2002. AVSS/NT also raises the possibility of making AVSS available on the Internet, which might be a better long-term solution to providing support to hundreds of systems throughout the state. Ron Williams proposed a 'central server' Internet model, actually composed of three separate servers at the state level: one for hospitals, one for LRDs, and one for OVR. This would require some significant changes to AVSS, for example, adding user name in addition to password for system access.

12. AVSS Technical Assistance.

1. Nearly all LRDs have paid their FY 99-00 contributions of $1,500 plus $500 per added AVSS site. Gail Cayton has written a Local Registrar's Manual, which was circulated and is available on the AVSS web site.

Acknowledgment: This summary is based on notes taken by Peter Chen, Emily Lehman, and Larry Portigal.

Updated June 7, 2000 by RL Williams
Return To AVSS Home Page