New Information for Policymakers: National ...

Section on Survey Research Methods – JSM 2009

New Information for Policymakers: National Ambulatory Medical Care Survey and National Hospital Ambulatory Medical Care Survey Esther Hing, M.P.H.1, Linda McCaig, M.P.H.1, Margaret Hall, Ph.D.1 1

National Center for Health Statistics, 3311 Toledo Road, Room 3409, Hyattsville, MD 20782

Abstract The National Center for Health Statistics, part of the Centers for Disease Control and Prevention, conducts two surveys of ambulatory care provided in the United States: the National Ambulatory Medical Care Survey (NAMCS) and the National Hospital Ambulatory Medical Care Survey (NHAMCS). NAMCS is a nationally representative survey of ambulatory visits to non-federal office-based physicians, while NHAMCS is its counterpart in emergency and outpatient departments of non-federal, short-stay or general hospitals. This paper highlights changes made to the surveys between 2004 and 2010. These include dual frame samples of community health center (CHC) physicians starting in 2006, dual mode surveys of physicians in 2008 and 2009, expansion of NHAMCS’s scope in 2009 to include ambulatory surgery centers previously included in the National Survey of Ambulatory Surgery, a cervical cancer screening supplement to both NAMCS and NHAMCS, supplements on both hospital pandemic and emergency response preparedness and availability of emergency pediatric services and equipment, and expansion of NHAMCS to measure correlates of emergency department crowding. These changes reflect the flexibility of NAMCS and NHAMCS to incorporate survey changes that improve the policy relevance of statistics, measure the diffusion of new technologies, and provide new data on our health care delivery system. Key Words: Ambulatory care provider surveys, Redesigned complex surveys 1. Introduction and background NAMCS and NHAMCS, conducted by the Centers for Disease Control and Prevention’s National Center for Health Statistics, have been measuring health care utilization in physician offices since 1973 and in hospital emergency and outpatient department utilization since 1992. Data from the two surveys serve a variety of analytic purposes. Survey content has been flexible, allowing changes to accommodate experts and other federal agencies in their quest for statistics that inform policymakers on the changing health care system. This paper addresses how the surveys were changed to improve policy-relevant statistics. 1.1 NAMCS survey design NAMCS collects data on visits made in the United States to the offices of non-federally employed physicians (excluding those in the specialties of anesthesiology, radiology, and pathology) who were classified by the American Medical Association (AMA) and the American Osteopathic Association (AOA) as providing "office-based, patient care." It utilizes a multistage probability sample design. Eligible sample physicians who agree to participate in the survey complete Patient Record forms for a systematic random sample

4607


of approximately 30 office visits occurring during a randomly assigned 1-week reporting period. Approximately 25,000 visits are sampled from about 2,000 physician offices each year (1).

1.2 NHAMCS survey design NHAMCS collects data on patient visits to emergency and outpatient departments of nonFederal, short-stay or general hospitals. Short-stay hospitals are those with an average length of stay of less than 30 days, while general hospitals include those whose specialty is general (medical/surgical) or children's general. Federal hospitals, hospital units of institutions, and hospitals with less than six beds are ineligible for NHAMCS. It employs a four-stage probability sample design involving samples of geographic primary sampling units, hospitals with emergency departments (EDs) and/or outpatient departments (OPDs) within primary sampling units, and clinics within outpatient departments. The final sampling stage involves systematic random samples of outpatient clinic/emergency department visits during randomly assigned 4-week reporting periods. Each year, approximately 35,000 encounters are sampled respectively, both in emergency departments and outpatient departments of about 480 non-Federal, short-stay, and general hospitals (2,3).

2. Survey Design Changes

2.1 Survey supplements Since 2002, NAMCS and/or NHAMCS have fielded special supplements for emerging needs as well as new technologies including: hospital emergency preparedness, availability of emergency pediatric services and equipment, and cervical cancer screening.

2.1.1 Cervical Cancer Screening Supplement The Cervical Cancer Screening Supplement to NAMCS and the OPD component of NHAMCS is being conducted from 2006 to 2010 and is sponsored by the Centers for Disease Control and Prevention’s National Center for Chronic Disease Prevention and Health Promotion. Human papilloma virus (HPV) testing and the HPV vaccine, approved for routine vaccination of girls 11-12 years of age, are shifting the paradigm of cervical cancer screening and management. The purpose of this supplement is to monitor changes in cervical cancer screening practices that may result from the introduction of these new technologies. The questionnaire is self-administered and is given to physicians and OPD clinic directors who conduct cervical cancer screening.

2.1.2 Emergency Pediatric Services and Equipment Supplement The Emergency Pediatric Services and Equipment Supplement was conducted in 2002, 2003 (4), and 2006 and was sponsored by the Health Resources and Services Administration (HRSA). HRSA wanted to replicate an emergency pediatric services study conducted in 1998 using the National Electronic Injury Surveillance System. The purpose of the supplement was to assess how well hospital EDs are prepared to provide pediatric emergency services. The supplement contained a short set of questions related to hospital services followed by a listing of 131 pieces of equipment recommended by the American Academy of Pediatrics to be available in EDs. The questionnaire was selfadministered and given to the ED director.

4608


2.1.3 Pandemic and Emergency Response Preparedness Supplement The Pandemic and Emergency Response Preparedness Supplement was sponsored by the Department of Health and Human Services and conducted in 2008. The purpose of the supplement was to assess progress towards hospital preparedness for terrorism, mass casualty incidents, natural disasters, and epidemics. Information was collected on the content of the hospital’s emergency response plan, staff training, participation in mass casualty drills, and the hospital’s resources and capabilities. This supplement was a modification of the “Bioterrorism and Mass Casualty Preparedness Supplement”, conducted in 2003 and 2004 (5), and was expanded to support the Department’s goal of preparing for emerging health threats, such as pandemic flu. The questionnaire was selfadministered and given to the person responsible for the hospital’s emergency response plan. 2.2 Expansion of Emergency Department questionnaire items to evaluate patient flow and crowding and to assess quality of care. A 2006 Institute of Medicine (IOM) report on the “Future of Emergency Care” cited NHAMCS data over 100 times (6). This report identified certain areas in which more data are needed to evaluate ED patient flow and issues related to ED crowding. This section discusses changes made to the ED questionnaire to address these emerging policy needs.

2.2.1 Induction interview additions In 2007, the NHAMCS induction form was modified based on the IOM report to determine if EDs were using the suggested strategies to reduce crowding. Because a lack of inpatient beds may be a contributor to ED crowding, questions were added at the hospital level about: • How many days in a week are inpatient elective surgeries scheduled • Does the hospital has a bed coordinator or czar • How often are hospital bed census data available. At the ED level, the following questions were asked of the administrator: • Does the ED have a physically separate observation or clinical decision unit o If yes, do ED physicians make decisions for patients in this unit • Are admitted patients ever “boarded” for more than 2 hours in the ED or observation unit while waiting for an inpatient bed • If the ED is critically overloaded are admitted patients ever “boarded” in inpatient hallways or in another space outside the ED • How many hours was the ED on ambulance diversion in the previous year • Is ambulance diversion managed on a regional or hospital level • Does the hospital continue to admit elective surgery cases when the ED is on ambulance diversion • How many standard treatment spaces and other treatment spaces does the ED have • Has the ED increased the number of treatment spaces in the last two years • Has the ED’s physical space been expanded in the last 2 years • Are there any plans to expand the ED’s physical space within the next two years. The ED administrator was also asked about strategies to optimize ED efficiency: • Bedside registration • Computer-assisted triage

4609


• • • • • • •

Separate fast track unit for nonurgent care Separate operating room dedicated to ED patients Electronic dashboard (i.e., displays updated patient information and integrates multiple data sources) Radio frequency identification tracking (i.e., shows exact location of patients, caregivers, and equipment) Zone nursing (i.e., all of a nurse’s patients are located in one area) “Pool” nurses (i.e., nurses that can be pulled to the ED to respond to surges in demand) Full capacity protocol (i.e., allows some admitted patients to move from the ED to inpatient corridors while awaiting a bed).

Ever board 90

Do not board 87

83

P e rc e n t d is trib u tio n

80 70

63

61

60 50 40

35

39

30 20

12

10

8

0 Total

Small

50,000

SOURCE: CDC/NCHS, National Hospital Ambulatory Medical Care Survey.

Figure 1: Percent distribution of EDs that ever “board” admitted patients for more than 2 hours by ED visit volume: United States, 2007

Figure 1 presents 2007 survey data on ED crowding. Sixty-three percent of all EDs boarded admitted patients for more than 2 hours while waiting for an inpatient bed. Eighty-three percent of EDs with a medium visit volume and 87% of EDs with a large visit volume boarded patients compared with 39% of EDs with the lowest visit volume (7).

2.2.2 New ED encounter items collected Multiple items related to quality of care and ED crowding were added to the ED Patient Record form from 2004 through 2010. Items include: • Dates when seen by a provider and ED discharge to validate that a patient’s length of stay in the ED was greater than 24 hours and to improve the quality of

4610


the data on waiting and treatment times (i.e., to validate that the patient was discharged on the day after ED arrival) • Patient residence (e.g., home, nursing home, homeless) • Respiratory rate • Pulse oximetry value which is a measure of oxygen saturation in the blood • Whether the patient is on oxygen on arrival so the pulse oximetry value can be interpreted • Glasgow Coma Scale which provides an assessment of the patient’s alertness. To assess previous care as well as quality of care, the following items were added: • Whether the patient was discharged from any hospital in the last 7 days • Number of times the patient was seen in the ED in the last 12 months. An increase in the number of patients with chronic diseases seeking care in the ED could contribute to crowding. In order to monitor these visits, a chronic disease checklist was added which includes the following diagnoses: • Cerebrovascular disease/history of stroke • Congestive heart failure • Condition requiring dialysis • HIV • Diabetes. The medication item now indicates whether the drug was given in the ED or prescribed at discharge. In an attempt to discern the intensity of resources utilized at an ED visit, service level as indicated by the Current Procedural Terminology code used for billing purposes was added. A distinction can now be made between ED patients who were discharged from the observation unit and those who were admitted to the hospital. • The date and time of discharge from an observation unit to home was added so the duration of observation stays can be calculated. • To address the amount of time a patient is boarded in the ED, dates and times were added to indicate when an inpatient bed was requested and when the patient left the ED. In 2005, additional information on patients admitted to the hospital was added to the ED Patient Record form: • Hospital unit the patient was admitted to (e.g., critical care, stepdown, detox) • Whether the patient was admitted by a hospitalist • Hospital discharge date • Principal discharge diagnosis • Hospital discharge status (i.e., whether discharged dead or alive) • Disposition (i.e., discharged to home, nursing home, or transferred to another facility). 2.3 Adoption of electronic health records reported in NAMCS and NHAMCS Since 2005, NAMCS and NHAMCS have collected data on use or availability of electronic health record (EHR) systems in physician offices, and hospital emergency and outpatient departments. When the scope of NHAMCS was expanded in 2009 to include hospital-based ambulatory surgery centers (ASCs), the same data items were also collected in these facilities (discussed further in section 2.4.3). The impetus for interest in

4611


these estimates was the 2004 Health Information Technology Initiative, whose goal was to have near universal adoption of EHR systems by health care providers by 2014. This was followed by the 2009 American Recovery and Reinvestment Act or ARRA, also known as the Stimulus Package, which has provided billions of dollars to support the adoption and use of EHRs (8). The respondent in each setting is asked if they have a computerized system for the following: patient demographic information and if yes, does it include patient problem lists; orders for prescriptions and if yes, are there warnings of drug interactions or contraindications and are prescriptions sent electronically to the pharmacy; orders for tests and if yes, are orders sent electronically; viewing of lab results and if yes, are out of range levels highlighted; viewing of imaging results and if yes, are electronic images returned; clinical notes and if yes, do they include medical history and follow-up notes; reminders for guideline-based interventions and/or screening tests; and public health reporting and if yes, are notifiable diseases sent electronically. Additional questions related to EHR use include the year that the EHR system was bought or last upgraded, if the EHR system is certified by the Certification Commission for Healthcare Information Technology, and if there are plans to install a new EHR system or replace an existing one within the next 3 years. 2.4 Changes made to expand or improve the precision of survey estimates This section highlights NAMCS’s use of dual-frame sampling techniques to provide community health center utilization estimates, and dual-mode survey procedures to increase the sample size for estimates of physician adoption of health information technology. Expanding the scope of NHAMCS to include ambulatory surgery centers is also discussed.

2.4.1 Dual-frame sampling to improve community health center estimates While physicians working in community health centers (CHCs) have always been in scope for NAMCS, too few of them were sampled to provide reliable estimates of physicians working in CHCs and their visits. However, President Bush’s 2002 Federal Health Center Growth Initiative highlighted the need for better estimates in this setting (9). To improve estimates of CHC providers and their visits in NAMCS, dual-frame sampling methods were utilized. Starting in 2006, a separate stratum of CHCs was selected in addition to the AMA/AOA sample of physicians. The sample of CHCs included all federally qualified health centers (FQHCs) that met criteria to receive federal funds. Sampled FQHCs included federally funded section 330 grantees, look-alike health centers that qualified for federal funds but did not receive any, and urban Indian Health Service outpatient clinics. FQHC physicians were dually sampled because the AMA/AOA frames do not identify where physicians work. Thus, a CHC physician could potentially be selected twice, first from the AMA/AOA frame, and second form the sampled CHC. To avoid double counting these physicians, those selected from the AMA/AOA frame that were identified as working in CHCs during the survey were omitted from the final survey file. Thus, physician visits selected in CHCs involved an additional level of sampling compared to physician visits sampled from the AMA/AOA frame. Another difference with the CHC sample is the sampling of visits to non-physician clinicians (that is, physician assistants, nurse practitioners, and nurse midwives). These clinicians were

4612


included because of the high utilization of non-physician clinician services in CHCs compared with physician offices. In 2006-07, nearly ninety percent of CHCs employed non-physician clinicians, and about 29% of all CHC visits were to non-physician clinicians (10,11). During the same time period, only 53% of non-CHC physicians employed non-physician clinicians. Comparable visits to non-physician clinicians working in the practices of physicians selected from the AMA/AOA frame were not sampled.

2.4.2 Dual-mode surveys to increase sample size for detailed analysis of physician adoption of electronic health records Starting in 2005, NAMCS began monitoring physician adoption of EHR systems (12). To improve the ability to conduct detailed analysis of which physicians adopt EHRs, the NAMCS sample of physicians was increased in both 2008 and 2009 by adding a mail survey of 2,000 physicians to the 3,200 physicians selected annually for NAMCS. The annual NAMCS sample involves personal interviews with physicians prior to selection of a sample of visits during a random week of the year. The content of the mail questionnaire was comparable to questions asked in the personal interview. Funding for the increased sample size was provided by the Office of the National Coordinator for Health Information Technology. 45 Any EMR

Basic EHR system

40

Percent of physicians

35 30 25 20 15 10 5 0 2001

2002

2003

2004

2005

2006

2007

2008

NOTES: EMR is electronic medical record. EHR is electronic health record. 2001-2007 estimates are based on personal interview, 2008 estimates are based on mail survey. SOURCE: CDC/NCHS, National Ambulatory Medical Care Survey, 2001-2008

Figure 2: Percentage of office-based physicians using electronic medical records and basic electronic health record systems: United States, 2001-2008.

Figure 2 shows that NAMCS has been monitoring physician use of electronic

medical record (EMR) systems since 2001. Since 2005, NAMCS has included detailed questions to help determine whether electronic systems used by physicians

4613


could be classified as EHRs, based on criteria developed by health information technology experts (12,13). These detailed questions were administered in both NAMCS and NHAMCS induction interviews as described in section 2.3. The 2008 estimates

shown in Figure 2 are from the NAMCS mail survey. 2.4.3 Expanding scope of NHAMCS to incorporate ambulatory surgery centers The National Survey of Ambulatory Surgery or NSAS was conducted from 1994 to 1996 and again in 2006. Both hospital-based and freestanding ambulatory surgery centers (ASCs) were sampled, accounting for about 35 million visits annually with 57% being hospital-based (14). Since it proved difficult to obtain funding for an annual NSAS, the scope of NHAMCS was expanded in 2009 to include these facilities. Data are currently being collected from approximately 16,000 ambulatory surgery encounters, taking place in about 200 hospital-based ASCs. In 2010, about 200 freestanding ASCs will be added as a separate stratum. A total of 32,000 encounters from both hospital-based and freestanding ASCs are expected in 2010.

3. Conclusions This paper describes the continuing efforts of NAMCS and NHAMCS survey designers to produce policy-relevant statistics that measure the diffusion of new technologies and changes in care provided by our health care delivery system.

References 1. Cherry DK, Hing E, Woodwell DA, Rechtsteiner EA. National Ambulatory Medical Care Survey: 2006 Summary. National health statistics report; no 3. Hyattsville, MD: National Center for Health Statistics. 2008. 2. Pitts SR, Niska RW, Xu J, Burt CW. National Hospital Ambulatory Medical Care Survey: 2006 Emergency Department Summary. National health statistics report; no 7. Hyattsville, MD: National Center for Health Statistics. 2008. 3. Hing E, Hall MJ, Xu J. National Hospital Ambulatory Medical Care Survey: 2006 Outpatient Department Summary. National health statistics report; no 4. Hyattsville, MD: National Center for Health Statistics. 2008. 4. Middleton KR, Burt CW. Availability of Pediatric Services and Equipment in Emergency Departments: United States, 2002-03. Advance data from vital and health statistics; no 367. Hyattsville, MD: National Center for Health Statistics. 2006. 5.Niska RW, Burt CW. Emergency response planning in hospitals: United States, 2003-04. Advance data from vital and health statistics; no 391. Hyattsville, MD: National Center for Health Statistics. 2007. 6. Committee on the Future of Emergency Care in the United States Health System. Hospitalbased emergency care: at the breaking point. National Academies Press. Washington, DC. 2007. 7. McCaig LF, Xu J, Niska RW. Estimates of Emergency Department Capacity: United States, 2007. Health E-Stat. Hyattsville, Maryland: National Center for Health Statistics. 2009. 8. Blumenthal D. Stimulating the Adoption of Health Information Technology. NEJM. 360(15): 1477-1479. 2009. 9. Inglehart JK. Spreading the Safety Net – Obstacles to the Expansion of Community Health Centers. N Engl J Med 358(13):1321-23. 2008. 10. Hing E, Hooker R. Care Provided by Non-physician Clinicians in Community Health Centers: United States, 2006-07. NCHS data brief. Hyattsville, MD: National Center for Health Statistics. In review.

4614


11. Hing E. “National Data on Health Centers and Their Providers”, March 27, 2009 presentation at the National Association of Community Health Centers’ Policy & Issues Forum; unpublished estimates from the 2006-07 National Ambulatory Medical Care Survey. 12. Blumenthal D, DesRoches C, Donelan K, et al. Health Information Technology in the United States: The Information Base for Progress. Robert Wood Johnson Foundation. 2006. 13. DesRoches CM, Campbell EG, Rao SR, et al. Electronic Health Records in Ambulatory Care – A National Survey of Physicians. NEJM 359(1):50-60. July 19, 2008. 14. Cullen KA, Hall MJ, Golosinskiy A. Ambulatory Surgery in the United States, 2006. National health statistics reports; no 11. Revised. Hyattsville, MD: National Center for Health Statistics. 2009.

4615