1 / 30

Improving cause-of-fetal death data to strengthen perinatal mortality prevention efforts

Improving cause-of-fetal death data to strengthen perinatal mortality prevention efforts. Donna L. Hoyert, Ph.D. Donna Glenn Marian F. MacDorman, Ph.D. National Center for Health Statistics. Goal. To release fetal cause-of-death data nationally. Importance.

kenny
Download Presentation

Improving cause-of-fetal death data to strengthen perinatal mortality prevention efforts

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. Improving cause-of-fetal death data to strengthen perinatal mortality prevention efforts Donna L. Hoyert, Ph.D. Donna Glenn Marian F. MacDorman, Ph.D. National Center for Health Statistics

  2. Goal To release fetal cause-of-death data nationally

  3. Importance • Fetal mortality is an important public health issue • Increased interest at NIH, CDC, Stillbirth Alliance, US Congress, and others • Just considering fetal deaths 20+ weeks gestation, there were almost 26,000 fetal deaths in 2005:

  4. Fetal, infant, and maternal deaths for selected years: US, 1935-2005 Infant deaths • Number of infant deaths converging toward fetal • Both are many times larger than maternal deaths Fetal deaths of 20+ weeks gestation Maternal deaths

  5. Relative number of fetal and infant deaths by area, 2005 It’s not very unusual for areas to have more fetal than infant deaths

  6. Needs and responses • Need to improve completeness and quality • Responding to these needs and increased interest in fetal deaths, we have been updating you on the following efforts: • Quality control [VSCP 2008] • Induced terminations [NAPHSIS 2009] • Automated coding…

  7. Steps toward achieving goal: Release fetal cause-of-death data nationally • 1921 Began routine fetal mortality publication, not including cause • 1968 Began to release public use data, not including cause • 1989 NCHS produced 1st coding guidelines • 1992 NCHS began routinely receiving cause data from areas • 2003 cause of death section of standard report of fetal death changed substantially • 2006 Began routine NVSR on fetal and perinatal mortality • 2007 Included fetal data in VitalStats data • 2009 NCHS powerpoint on coding distributed to areas

  8. Steps toward achieving goal: Release fetal cause-of-death data nationally • 1921 Began routine fetal mortality publication, not including cause • 1968 Began to release public use data, not including cause • 1989 NCHS produced 1st coding guidelines • 1992 NCHS began routinely receiving cause data from areas • 2003 cause of death section of standard report of fetal death changed substantially • 2006 Began routine NVSR on fetal and perinatal mortality • 2007 Included fetal data in VitalStats data • 2009 NCHS powerpoint on coding distributed to areas Previous efforts not enough

  9. Problemsin context of new fetal cause format • Not all areas using the 2003 format • Not all areas using the 2003 format code cause of fetal death • As usual, certifiers enter information in all sorts of ways:

  10. Placement of fetal causes using 2003 format Incorrect: Other problems (e.g., multiple causes in 18a) Correct: Single cause in 18a, maybe additional info in 18b Incorrect: Single cause in 18b, no cause in 18a, blank Incorrect: Single cause in 18a, but repeated in 18b

  11. Other reporting characteristics • Use of checkboxes and specify lines to report cause • 67% reports have a checkbox marked • 69% reports have an entry in a specify line • 39% reports have both checkboxes and specify lines • Most common problem is mirror reporting (i.e. 18a is identical to 18b at 20%) • Average – 2 conditions

  12. Summary of current approach • Has not resulted in complete coded national data • Recent changes (e.g., revised report, resource constraints) pose more challenges • Inconsistency in how code:

  13. Comparing NCHS and State code agreement Note: No state codes for other 5 states.

  14. Is there an alternative approach? • Examine if there might be another approach • Feasibility study • Test project • Tentative exploration of data

  15. Feasibility study • 2009: contract to look into how to move forward • Decision: Develop incrementally • develop program to code the checkboxes and facilitate manual coding of literal entries • develop program to select initiating cause • review data and revise processing

  16. 1st stage: Test project • Mainly manually code at NCHS • Assign codes for checkboxes • Forced use of ACME • Workarounds: • P95 fetal death code is not valid for ACME. Used P969 to allow system to process. Changed to P95 • Ill-defined P20.1, P20.9, and P21.9 are not included in ACME ill-defined table • Codes in 18a entered on line 1 • Codes in 18b entered on Part II

  17. Assessment • Can increase the proportion of records with coded data available using alternative approach • Initial effort requires manual coding for around 70% of records, similar to feasibility study estimates • Can improve consistency of coding by about 20%

  18. Assessment • Need further decisions on some specific coding situations, requirements, develop tools, and document decisions: • Develop index for fetal death categories • Expand list of abbreviations for fetal deaths • Identify terms implying abortion • Develop dictionary to be used for automation

  19. Examples of NCHS coded data

  20. Selected initiating causes of fetal death Note: 20+ weeks gestation

  21. Selected initiating causes of fetal death by obstetrical estimate of gestation Note: 20+ weeks gestation Percent

  22. Percent mentioning congenital malformations as a cause by maternal age Note: 20+ weeks gestation Years of age

  23. Summary • An alternative approach would: • Ensure coding done consistently • Improve cause of fetal death data, and that, in turn, would strengthen perinatal mortality prevention efforts • Could better respond to increased interest in fetal death • Would have better data to be able to target prevention efforts at high-risk groups

  24. Summary of initial efforts • While have made forward strides towards goal of releasing cause over time, have not realized goal • Latest activities at NCHS: Feasibility study, initial steps towards developing automated system, and exploration of data • Can expand the number of areas with coded data • Can automate coding of 1/3rd of records with trivial effort • Expect can quickly expand the proportion can code automatically • Initial development tools match our expected codes reasonably well

  25. How to continue to develop coding project

  26. Now is still not the time to automate • Cost of full automation is not justified by the number of records • System requirements are not yet clearly defined • Continue manual coding until a system can be defined

  27. State actions • Provide education to certifiers • Enter appropriate conditions in each category • Maternal Conditions: congenital heart defect – is this maternal or fetus? • Do not use abbreviations AMA: advanced maternal age arthrogryposis multiplex congenital

  28. State actions • Add spell checker to data entry systems including electronic registration system

  29. NCHS actions • Develop complete fetal death coding instructions • Instructions for external causes • Instructions for maternal conditions • Develop a more complete index for fetal deaths • Update the valid code list

  30. Requirements for new system • Update valid code list • Changes in ACME modification tables • Remove “due to” linkages • Remove entries which have invalid fetal death codes • Create trivial table • Create ill-defined table • Review ICD for additional table entries

More Related