1. Using past visit information to enhance analysis of National Ambulatory Medical Care Survey (NAMCS) data Session 25 July 13, 2004 10:30-noon

2. Session overview  Introduction and methodology for using past visit information – Catharine Burt  Example of analysis of quality of care for hypertension – Esther Hing

3. Using past visit information to enhance analysis of ambulatory care from encounter data Catharine W. Burt, Ed.D. Chief, Ambulatory Care Statistics Branch July 13, 2004 Data Users Conference U.S. DEPARTMENT OF HEALTH AND HUMAN SERVICES Centers for Disease Control and Prevention National Center for Health Statistics

4. Topics  Introduction  Brief survey background  Past visit item  Re-weighting methodology  Analysis using re-weighting  Summary

5. Encounter vs. population-based surveys that measure ambulatory care utilization  Population-based National Health Interview Survey (NHIS) National Health Interview Survey (NHIS) National Health and Nutrition Examination Survey (NHANES) National Health and Nutrition Examination Survey (NHANES) Medical Expenditure Panel Survey (MEPS) Medical Expenditure Panel Survey (MEPS)  Encounter-based National Ambulatory Medical Care Survey (NAMCS) National Ambulatory Medical Care Survey (NAMCS) National Hospital Ambulatory Medical Care Survey (NHAMCS) National Hospital Ambulatory Medical Care Survey (NHAMCS)

6. Pattern of ambulatory care visits according to patient age: NHIS, 2001 <06 6-17 18-24 25-44 45-64 65-74 75+ 0%20%40%60%80%100% None1-3 visits4-9 visits10 or more visits

7. Rate of visits per population by patient age: NAMCS, NHAMCS-OPD, 2001 3.4 4.2 1.9 2.7 4.1 6.7 7.8 All ages <06 6-17 18-24 25-44 45-64 65-74 75+ Patient age in years 0246810 Rate

8. Desired goals from adding number of past visits to encounter data Describe patterns of care by frequency of visiting the doctor Describe patterns of care by frequency of visiting the doctor Provide more information about patients from encounter-level data Provide more information about patients from encounter-level data Better describe quality of care to patients vs. describing content of encounter Better describe quality of care to patients vs. describing content of encounter

9. NAMCS and NHAMCS background  NAMCS Fielded 1973-1981, 1985, 1989-present Fielded 1973-1981, 1985, 1989-present 3-stage sample 3-stage sample PSUs –PSUs – physicians – physicians – visits during 1 week visits during 1 week  NHAMCS Fielded annually since 1992 Fielded annually since 1992 4-stage sample 4-stage sample PSUs –PSUs – hospitals – ED/OPD clinics – visits during 4 weeks visits during 4 weeks SEs rates percents drugs providers visits sample records

10. Scope of the NAMCS  Physicians must be: Primarily engaged in office-based, patient care Primarily engaged in office-based, patient care Non-federally employed Non-federally employed Not in anesthesiology, radiology, or pathology Not in anesthesiology, radiology, or pathology

11. In-Scope Visits for NAMCS  Basic unit of sampling is the physician- patient visit  Visit must occur in a physician’s office  Visits must be for medical purposes  Out-of-scope contacts include: Visits for administrative purposes Visits for administrative purposes House calls, phone calls, emails, visits in hospital setting unless physician has private office there House calls, phone calls, emails, visits in hospital setting unless physician has private office there

12. In-Scope NAMCS Locations  Freestanding private solo or group practice  Freestanding clinic/urgicenter  Neighborhood medical and mental health centers  Privately operated clinics  Non-Federal government clinic  Health maintenance organization  Federally qualified health center  Faculty practice plan

13. Out-of-Scope NAMCS Locations  Hospital ED’s and OPD’s  Ambulatory surgicenter  Institutional setting (schools, prisons)  Industrial outpatient facility  Federally operated clinic  Laser vision surgery

14. Scope of the NHAMCS  Basic unit of sampling is patient visit  Emergency and outpatient departments of non-federal, general and short-stay hospitals  Not Federal, military, or Veterans Administration facilities  Located in 50 states and D.C.

15. Items Collected  Patient characteristics Age, sex, race, ethnicity Age, sex, race, ethnicity  Visit characteristics Source of payment, continuity of care, reason for visit, diagnosis, treatment, medications ordered or provided Source of payment, continuity of care, reason for visit, diagnosis, treatment, medications ordered or provided  Provider characteristics Physician specialty, hospital ownership Physician specialty, hospital ownership  Drug characteristics Therapeutic class, composition, control status, ingredients, Rx or OTC Therapeutic class, composition, control status, ingredients, Rx or OTC

18. Percent distribution for number of past visits within l2 months: NAMCS, 2001 New patient 01-23-56+Unknown Number of past visits 0 5 10 15 20 25 30 Percent of visits

19. Percent distribution of past visit item by PCP status: NAMCS 2001 New patient 01-23-56+ Unknown Number of past visits 0 5 10 15 20 25 30 35 Percent of visits PCPOther

20. Percent of visits where the patient made 6 or more visits in past year by patient age: NAMCS 2001

21. Percent of visits to patient's primary care provider (PCP) by number of past visits within last l2 months: NAMCS, 2001 New patients 01-23-56+ Number of past visits 0 20 40 60 80 100 Percent of visits

22. Drug mention rate by past visits: NAMCS, 2001 New patient 01-23-56+ Number of past visits 0 50 100 150 200 Number of drugs per 100 visits

23. Drug mention rates by number of past visits for selected therapeutic classes: NAMCS, 2001 New Patient 01-23-56+ Number of past visits 0 5 10 15 20 25 30 35 Number of drugs per 100 visits Cardiovascular CNS

24. Drug mention rates by number of past visits for selected therapeutic classes: NAMCS, 2001 New Patient 01-23-56+ Number of past visits 0 5 10 15 20 25 30 35 Number of drugs per 100 visits Antimicrobials Cardiovascular CNS Pain relief

25. Re-weighting methodology  Assumptions and definitions Dividing the visit weight by number of visits within year provides a measure of patients Dividing the visit weight by number of visits within year provides a measure of patients Patient is relation between person and sampled doctor Patient is relation between person and sampled doctor One person can be different patients to different doctors One person can be different patients to different doctors Assumes previous visits by same patient have similar visit characteristics Assumes previous visits by same patient have similar visit characteristics

26. 3 patients making a total of 7 visits

27. 4/7 2/71/7 Probability of selecting a visit

28. 4/72/7 1/7 1.5.25 Multiplicity factor

29. Creation of a re-weighting factor Item categories Past visits Annual visits Interval midpoint Viz ratio New0111 00111 1-21-22-32.5.4 3-53-54-65.2 6+6+7+8.125

30.  Patient weight = Visit weight * Viz ratio  Visit weight → 880,487,000 visits ± 33,373,000 ± 33,373,000  Patient weight → 322,739,000 patients ± 13,073,000 ± 13,073,000

31. Re-weighting example Annual visits Interval midpoint Visit weight Viz ratio Patient weight 1110110 2-32.520.48 4-6540.28 7+840.1255 Sum = 110 visits 31 patients

32. Analysis using re-weighting  Comparison of visits and patients  Analysis of patients

33. Comparison of distributions for visits and patients: NAMCS 2001 VisitsPatients 12-34-67+12-34-67+ 0 10 20 30 40 50 60 Percent

34. Percent distribution for people making any health care visits by number of visits made in one year: NHIS, 1999-2000 1-34-910+ Number of visits 0 10 20 30 40 50 60 Percent of persons Rate of persons making no health care visit was 17.5.

35. Rate of visits per population by patient age: NAMCS, NHAMCS-OPD, 2001 3.4 4.2 1.9 2.7 4.1 6.7 7.8 All ages <06 6-17 18-24 25-44 45-64 65-74 75+ Patient age in years 0246810 Rate

36. Components of utilization rates Rate of visits per population = Rate of visits per patient multiplied by multiplied by Rate of patients per population

37. Rate of visits per patient and patients per population by patient age: NAMCS, NHAMCS-OPD, 2001 All ages <06 6-17 18-24 25-44 45-64 65-74 75+ Patient age in years 00.511.522.533.5 Rate Visits per patient Patients per pop

38. Comparison of population-based and encounter-based estimates of ambulatory care utilization by patient age: NAMCS, NHAMCS-OPD, NHIS 2001 Source:NHIS,NAMCS & NHAMCS-OPD <066-1718-2425-4445-6465-7475+ Patient age 0 0.2 0.4 0.6 0.8 1 Probability 0 2 4 6 8 10 Rate per pop 4+ visits Any visits Patients Visits

39. Percent of visits and patients where the patient made 7 or more visits in the year by patient age: NAMCS 2001 VisitsPatients

40. Percent of visits and patients where the patient made only 1 visit in the year by patient age: NAMCS 2001 VisitsPatients

41. Rate per 100 pop of patients to same provider by type of setting and patient age: NAMCS, NHAMCS-OPD, 2001 <15 15-24 25-44 45-64 65-74 75+ Patient age in years 050100150200250 Number of patients per 100 pop Primary careSurgical specMedical specOPD

42. Rate per 100 pop of patients by number of visits within a year to same provider and PCP status: NAMCS, 2001 1 2-3 4-6 7+ Number of visits 0102030405060 Number of patients per 100 pop PCPNon-PCP

43. Rate per 100 pop of patients by number of visits within a year to PCP: NAMCS and NHAMCS-OPD, 2001 Private ins Medicaid Uninsured 01020304050 Number of patients per 100 pop 12-34-67+

44. Incidence of tobacco use cessation counseling at PCP visits: NAMCS & NHAMCS-OPD, 2001 23.7 27.3 VisitsPatients 0 5 10 15 20 25 30 Percent

45. Percent of visits by tobacco users to their PCP with tobacco cessation counseling by number of past visits within l2 months: NAMCS & NHAMCS-OPD, 2001 New patient01-23-56+ Number of past visits 0 10 20 30 40 Percent of visits

46. Public health example  Women at risk for bearing children with birth defects Levothyroxine (evidence of hypothyroid) Levothyroxine (evidence of hypothyroid) 2.6 million visits 2.6 million visits 2.3 million patients 2.3 million patients Accutane Accutane 397 thousand visits397 thousand visits 272 thousand patients272 thousand patients

47. Comparison of drug mention rates for visits and patients Drug class Visit DMR Patient DMR Ratio CNS13101.35 Cardio22171.29 Pain relief 18171.10 Antimicrobial13121.08 DMR is number of drugs per 100 visits/patients.

48. Ratio of drug mention rates based on visits and patients: NAMCS, 2001,,,, CNSCardioPain relief Antimicrobial. 0.6 0.8 1 1.2 1.4 1.6 1.8

49. Summary  Past visits items provides depth to analysis of ambulatory care utilization  Visit records may be re-weighted to provide patient-level estimates  Re-weighted distribution more closely resembles population-based estimates  No change in sampling variance estimation procedure other than using the new weight  Caveats Assumption of similar characteristics is not applicable to all analytical variables Assumption of similar characteristics is not applicable to all analytical variables Still is not equivalent to person-level estimates Still is not equivalent to person-level estimates

50. Information available on the Internet  NCHS website is www.cdc.gov/nchs www.cdc.gov/nchs  National Health Care Survey website Ambulatory Health Care Data (AHCD) Ambulatory Health Care Data (AHCD)  Join our LISTSERV