Chiropractors as Safety Net Providers: First Report of Findings and Methods from a US Survey of...

11
CHIROPRACTORS AS SAFETY NET PROVIDERS: FIRST REPORT OF FINDINGS AND METHODS FROM A US SURVEY OF CHIROPRACTORS Monica Smith, DC, PhD, a and Lynne A. Carber, MPM b ABSTRACT Objective: This study evaluates the actual or potential contribution of the chiropractic profession in meeting US healthcare workforce needs. Methods: The authors performed a descriptive cross-sectional mail survey of US chiropractors in 2002 to 2003. Results: The amount of charity care provided by chiropractors closely approximates that of medical physicians; on a weekly basis, approximately 2 weekly hours of chiropractic care are provided for free, and 4 weekly hours are provided at a reduced fee. Conclusion: Chiropractors serve a vital, though often overlooked, role in the US health care safety net. (J Manipulative Physiol Ther 2007;30:718-728) Key Indexing Terms: Chiropractic; Uncompensated Care; Survey methods; Demographics C oupled with growing interest in the role of chiropractic in US healthcare delivery, recent policy mandates have accelerated the chiropractic health services research agenda. The Health Care Safety Net Improvement Act (HR 3450) includes a provision to make doctors of chiropractic (DCs) eligible to participate in the National Health Service Corps program. This project surveyed US chiropractors to gather information about the actual or potential contribution of the chiropractic profession in meeting healthcare workforce needs. This first published report of findings from this survey of chiropractors also includes a detailed description of the complex survey methods used in this project. METHODS Sampling Frame This project is a descriptive cross-sectional study, using a set of mail surveys distributed in 2002 to 2003 to 3 separate samples of US chiropractors (Fig 1). The first sample was a simple random sample of 2500 DCs nationwide, hereafter called simple random sample(SRS). The second sample was a clustered random sample of 2531 DCs that was drawn from the sites of the Community Tracking Study (CTS), hereafter called CTS sample.The third sample, a stratified random health professional shortage areas (HPSA) sample of 900 DCs, purposively oversampled chiropractors from medically under- served areas (HPSA). The samples were drawn consecutively (SRS first, CTS second, HPSA last, nonreplacement) from a list of 67,217 actively licensed DCs that was compiled from all US chiropractic state licensing boards in 2001. The CTS is conducted by the Center for Studying Health System Change bWeb site: http://www.hschange. comN to document changes in health care systems over time and to track the effect of those changes on people by describing and analyzing how the interactions of providers, insurers, policy makers, and others determine the accessi- bility, cost, and quality of locally delivered health care. The CTS is a longitudinal study and includes periodic national surveys of households (including an insurer followback), physicians, and employers, with survey samples randomly selected to provide a representative profile of change across the United States. The randomly selected sites of the CTS, mapped to chiropractic college locations, are portrayed in Figure 2. The HPSA-stratified sampling frame for this survey of chiropractors was defined using the 1996 Area Resource File, a database maintained by the Health Resources and Services Administration Bureau of Health Professions that contains county-level information on health facilities, health profes- sions, measures of resource scarcity, health status, economic activity, health training programs, and socioeconomic and a Associate Professor, Palmer Center for Chiropractic Research, Davenport, Iava. b Research Associate II, Palmer Center for Chiropractic Research, Davenport, Iava. Support for this project was provided through the Council on Chiropractic Guidelines and Practice Parameters and Palmer Center for Chiropractic Research. The authors report no conflicts of interest. Submit requests for reprints to: Monica Smith, DC, PhD, Research fellow, UCSF-IHPS, 3333 California Street, Suite 265, San Francisco, CA 94118 (e-mail: [email protected]). Paper submitted April 26, 2007; in revised form September 7, 2007; accepted November 9, 2007. 0161-4754/$32.00 Copyright © 2007 by National University of Health Sciences. doi:10.1016/j.jmpt.2007.11.001 718

Transcript of Chiropractors as Safety Net Providers: First Report of Findings and Methods from a US Survey of...

Page 1: Chiropractors as Safety Net Providers: First Report of Findings and Methods from a US Survey of Chiropractors

CHIROPRACTORS AS SAFETY NET PROVIDERS:FIRST REPORT OF FINDINGS AND METHODS

FROM A US SURVEY OF CHIROPRACTORS

Monica Smith, DC, PhD,a and Lynne A. Carber, MPMb

a Associate ProDavenport, Iava.

b Research AsResearch, Davenpo

Support for thiChiropractic Guidefor Chiropractic Re

Submit requestsfellow, UCSF-IHPSCA 94118 (e-mail:

Paper submitte2007; accepted No

0161-4754/$32Copyright © 20doi:10.1016/j.jm

718

ABSTRACT

Objective: This study evaluates the actual or potential contribution of the chiropractic profession in meeting UShealthcare workforce needs.Methods: The authors performed a descriptive cross-sectional mail survey of US chiropractors in 2002 to 2003.Results: The amount of charity care provided by chiropractors closely approximates that of medical physicians; on aweekly basis, approximately 2 weekly hours of chiropractic care are provided for free, and 4 weekly hours are providedat a reduced fee.Conclusion: Chiropractors serve a vital, though often overlooked, role in the US health care safety net. (J ManipulativePhysiol Ther 2007;30:718-728)

Key Indexing Terms: Chiropractic; Uncompensated Care; Survey methods; Demographics

Coupled with growing interest in the role ofchiropractic in US healthcare delivery, recent policymandates have accelerated the chiropractic health

services research agenda. The Health Care Safety NetImprovement Act (HR 3450) includes a provision to makedoctors of chiropractic (DCs) eligible to participate in theNational Health Service Corps program. This projectsurveyed US chiropractors to gather information about theactual or potential contribution of the chiropractic professionin meeting healthcare workforce needs. This first publishedreport of findings from this survey of chiropractors alsoincludes a detailed description of the complex surveymethods used in this project.

METHODS

Sampling FrameThis project is a descriptive cross-sectional study, using a

set of mail surveys distributed in 2002 to 2003 to 3 separate

fessor, Palmer Center for Chiropractic Research,

sociate II, Palmer Center for Chiropracticrt, Iava.s project was provided through the Council onlines and Practice Parameters and Palmer Centersearch. The authors report no conflicts of interest.for reprints to:Monica Smith, DC, PhD,Research, 3333California Street, Suite 265, San Francisco,[email protected]).d April 26, 2007; in revised form September 7,vember 9, 2007..0007 by National University of Health Sciences.pt.2007.11.001

samples of US chiropractors (Fig 1). The first sample was asimple random sample of 2500 DCs nationwide, hereaftercalled “simple random sample” (SRS). The second samplewasa clustered random sample of 2531 DCs that was drawn fromthe sites of the Community Tracking Study (CTS), hereaftercalled “CTS sample.” The third sample, a stratified randomhealth professional shortage areas (HPSA) sample of 900DCs,purposively oversampled chiropractors frommedically under-served areas (HPSA). The samples were drawn consecutively(SRS first, CTS second, HPSA last, nonreplacement) from alist of 67,217 actively licensedDCs that was compiled from allUS chiropractic state licensing boards in 2001.

The CTS is conducted by the Center for StudyingHealth System Change bWeb site: http://www.hschange.comN to document changes in health care systems overtime and to track the effect of those changes on people bydescribing and analyzing how the interactions of providers,insurers, policy makers, and others determine the accessi-bility, cost, and quality of locally delivered health care.The CTS is a longitudinal study and includes periodicnational surveys of households (including an insurerfollowback), physicians, and employers, with surveysamples randomly selected to provide a representativeprofile of change across the United States. The randomlyselected sites of the CTS, mapped to chiropractic collegelocations, are portrayed in Figure 2.

The HPSA-stratified sampling frame for this survey ofchiropractors was defined using the 1996 Area Resource File,a database maintained by the Health Resources and ServicesAdministration Bureau of Health Professions that containscounty-level information on health facilities, health profes-sions, measures of resource scarcity, health status, economicactivity, health training programs, and socioeconomic and

Page 2: Chiropractors as Safety Net Providers: First Report of Findings and Methods from a US Survey of Chiropractors

Fig 1. Non-replacement sampling frame for set of 3 survey cohorts.

719Smith and CarberJournal of Manipulative and Physiological TherapeuticsChiropractors as Safety Net ProvidersVolume 30, Number 9

environmental characteristics bWeb site: http://bhpr.hrsa.gov/healthworkforceN. The distribution of US chiropractors,mapped to Health Profession Shortage Areas and chiropracticcollege locations, has been reported elsewhere.1

Survey ItemsThe survey questions had been previously validated by

pilot testing on a sample of 104 “key informants” drawnfrom the leadership rosters of the Congress of ChiropracticState Associations, the Federation of Chiropractic LicensingBoards, and the National Board of Chiropractic Examiners.

Basic demographic information was queried of all surveyparticipants for all 3 samples, as was information about theirchiropractic practice market service area (the area fromwhich they drew patients) and information about theirpractice volume and patient wait times.

In addition to the general questions posed to all surveyparticipants, subsets of survey respondents also were queriedon additional specific survey items. For instance, a subset ofthe surveyed chiropractors were asked about their provisionof uncompensated charity care, defined as that care providedat either no fee or reduced fee because of the financial needof the patient. These queries replicated survey items from thephysician survey of the CTS, which were designed to gatherinformation about the role of medical physicians in thenation's health care “safety net” and their provision ofuncompensated charity care in private practice.

Examples of some of the categorical and open-endeditems as queried on various versions of the survey ofchiropractors are provided in Appendix A.

Survey Data Collection and Data ManagementThe first round of surveys was mailed to all sampled

chiropractors in May 2002, with a follow-up mailing of ashorter version of the survey to nonrespondents in June toJuly 2002 and a second follow-up mailing of the shortersurvey in August to September 2002. Each survey mailingincluded a cover letter that emphasized the survey'simportance and assured respondents of their anonymityand a postage-paid return envelope. Each of these surveytechniques improved our response rate cost-effectively.2-6

For surveys returned as undeliverable (United States PostalService [USPS] Bad Address), we attempted to locate thesenonrespondents if possible and to reinitiate the 3-mailingprotocol for found nonrespondents. After 3 unreturnedmailings, we used extensive Internet-based search protocolsto locate telephone contact information for nonrespondents ifpossible. Nonrespondents for which we found a phone numberwere followed up by phone from October 2002 throughJanuary 2003 to encourage greater compliance, to assesssystematic bias in nonrespondents, and to calculate a moreaccurate effective response rate by adjusting for DCs listed asState Board licensed but not practicing. If we phone-verifiedthe contact information as correct (eg, confirmation by theDC's office staff or their answering machine message) butwere unable to directly contact the DC after multiple attempts,then the DC was classified as a found nonrespondent.Nonrespondents were classified as invalid if we could notfind them (eg, USPS BadAddress, NoWhite Page, NoYellowPage listing), or if we could not verify their contact information(eg, no answer after multiple phone attempts). Data algorithmsfor each of the 3 survey samples (SRS, CTS, HPSA samples)are detailed in Appendix B.

Categorical data were key double-entry–verified. Open-ended narrative response data were single-entered verbatiminto the database to support future qualitative and indepthsecondary analyses of these data, for example, allowing laterrecoding or categorization of narrative responses. Electronicdata were managed in password-protected relational data-bases on a secure network; hardcopy data were stored inlocked cabinets in the program offices during data entry andthen archived in a secured location. Databases were furthermanaged using Microsoft Access 2003 (Microsoft, Red-mond, WA) or SAS for Windows (Release 8.0) (SASInstitute Inc, Cary, NC), as needed. Data analyses reportedbelow were conducted using SPSS for Windows (Version12.0) (SPSS Inc, Chicago, Ill).

RESULTS

The separate and distinct sampling frames used for thissurvey requires that each set of data be analyzed and reportedseparately. For response rates and general demographics,results are reported for all 3 survey samples: SRS, CTS,and HPSA.

Page 3: Chiropractors as Safety Net Providers: First Report of Findings and Methods from a US Survey of Chiropractors

Fig 2. Chiropractic college locations mapped to randomly selected sites of the Community Tracking Study conducted by the Center forStudying Health System Change.

Table 1. Effective survey response rates, each sample

SampleMail survey effectiveresponse rate ⁎

Mail and phone successfulcontact rate ⁎

SRS 53.8% 58.9%CTS 49.7% 54.5%

720 Journal of Manipulative and Physiological TherapeuticsSmith and CarberNovember/December 2007Chiropractors as Safety Net Providers

The queries on provision of uncompensated charity carewere posed to all 3 samples for only the original (long-form)mailing. For the 2 follow-up mailings (short form), thequeries on uncompensated care were posed only to the CTSsample. Therefore, the results on uncompensated care arereported for the entire CTS sample and for a subsample of theSRS and HPSA groups (those that responded to the originallong-form mail survey).

Response Rate and Potential Nonresponse BiasEffective response rates for each sample are reported in

Table 1.We computed effective response rates as described inAppendix B and classified as “invalid” those nonrespondentswho were not found, not practicing, retired, or deceased.Overall, the response to our mailed surveys (Mail SurveyReturned Completed/Valid Mail Surveys Sent) was approxi-mately 50%. Including the phone follow-up to compute a rateof successful contact (Mail Survey Returned Completed plusPhone Follow-Up Completed/Valid Mail Surveys Sent)yielded an overall contact rate of approximately 60%.

Our phone follow-up to mail survey nonrespondents wasbrief and queried nonrespondents on their basic demo-

graphic information. Comparison of nonrespondent torespondent demographics are reported in Table 2. Therewere no significant differences between respondents andnonrespondents on practice status (full-time vs part-time)on whether the DC had more than 1 practice location or onwhether the DC anticipated being in the same practicelocation for more than 2 years. Compared with mail surveyrespondents, nonrespondents contacted during our phonefollow-up were significantly older and in practice longerand were more likely to be in solo practice rather thangroup and to not belong to a national or state professionalassociation. Of borderline significance (P = .07), mailsurvey respondents were slightly more likely to havegraduated from Logan, Life West, Cleveland Los Angeles,

Page 4: Chiropractors as Safety Net Providers: First Report of Findings and Methods from a US Survey of Chiropractors

Table 2. Comparison of mail survey respondents to nonrespondents

Respondents Nonrespondents

Full-time practice 85% 80%

N1 Practice location 12% 14%

Solo practice ⁎ 69% 78%

Plan on practicing in same location in 2 y 84% 85%

Direct patient care hours per week, main practice (mean [SD] ⁎⁎ 33.4 (10.7) 30.7 (11.8)

Total main practice hours per week, includes direct patient careplus all other hours spent in main practice (mean [SD])

36.1 (13.3) 35.9 (16.3)

Belong to national or state professional association ⁎⁎⁎ 64% 45%

Age (mean [SD]) ⁎ 46.5 (76.9) 62.2 (167.4)

Years since graduation from chiropractic college (mean (SD)] ⁎⁎ 16.5 (55.4) 33.5 (169.8)

Chiropractic college (%)

Cleveland (Los Angeles) 3.2 0.7

Cleveland (Kansas City) 3.0 3.1

Cleveland (unspecified campus) 0.5 0.7

Life (Georgia) 2.2 5.5

Life (West) 3.3 0.7

Life (unspecified campus) 8.5 11.7

Logan 9.4 7.8

Los Angeles (now Southern California University of Health Sciences) 5.5 7.0

National 9.2 8.6

New York 7.4 8.6

Northwestern 6.3 7.0

Palmer (Iowa) 6.5 10.9

Palmer (West) 3.4 3.1

Palmer (unspecified campus) 18.1 8.6

Parker 3.1 1.6

Sherman 2.1 4.7

Texas 3.7 6.3

Bridgeport 0.4 –

Western states 3.8 3.1

Other non-US or chiropractic colleges not listed above 0.4 0.3

⁎ P ≤ .05, significant difference between respondents and nonrespondents.⁎⁎ P ≤ .01, significant difference between respondents and nonrespondents.⁎⁎⁎ P ≤ .001, significant difference between respondents and nonrespondents.

721Smith and CarberJournal of Manipulative and Physiological TherapeuticsChiropractors as Safety Net ProvidersVolume 30, Number 9

Palmer (unspecified campus), and Parker chiropracticcolleges, whereas nonrespondents were more likely gradu-ates of Sherman, Life Georgia, Life (unspecified campus),Los Angeles, and Texas chiropractic colleges. Comparingrespondents to nonrespondents on their reported time spent inpractice per week, direct patient care hours were slightly yetsignificantly higher for respondents, but there were nosignificant differences for all practice hours (total patient careand non–patient care hours).

During the phone follow-up, we also asked nonrespon-dents their reason for not returning the mail survey; the most

common reasons given were that they were too busy (43%)or that they do not do surveys (24%). The phone follow-upalso asked the chiropractor's opinion on various chiropracticpractice guidelines (reported elsewhere); whether theirchiropractic practice was “cash-only” (92% replied “no”);and whether they had obtained student loans, the amountborrowed, and the amount outstanding. Of those phone-contacted chiropractors who had obtained student loans(71% of our phone survey participants), the average amountborrowed was $42 000, and 84% of those who had borrowedstill had an outstanding balance owed that averaged $14 000.

Page 5: Chiropractors as Safety Net Providers: First Report of Findings and Methods from a US Survey of Chiropractors

Table 4. Chiropractor provision of uncompensated safety net care

SRS CTS HPSA

DCs report zero weekly hours provided at no fee 30% 26% 30%

DCs report zero weekly hours provided at reduced fee 20% 14% 28%

DCs report zero weekly patient visits provided at no fee 29% 26% 27%

DCs report zero weekly patient visits provided at reduced fee 18% 15% 26%

For DCs who do provide uncompensated care:

No. of weekly hours provided for no fee (mean [SD]) 1.9 (1.6) 2.2 (2.5) 2.0 (2.0)

No. of weekly hours provided for reduced fee: (Mean [SD]) 4.0 (4.6) 4.8 (5.7) 3.7 (4.4)

No. of weekly patient visits provided for no fee (Mean [SD]) 6.2 (5.8) 5.6 (4.6) 5.7 (4.7)

No. of weekly patient visits provided for reduced fee (Mean [SD]) 15.0 (24.6) 17.4 (29.5) 13.4 (14.5)

Percentage of patient care hours provided for no fee 5.9% 6.7% 6.0%

Percentage of patient care hours provided for reduced fee 11.8% 14.4% 11.3%

Percentage of patient visits provided for no fee 6.8% 7.2% 6.2%

Table 3. Chiropractor demographics

SRS Sample CTS Sample HPSA Sample

Male 80% 75% 85%

White 92% 91% 96%

Age (mean [SD]) 46.9 (80.6) 47.1 (88.5) 44.1 (11.2)

25-35 24% 25% 26%

36-45 35% 34% 32%

46-55 29% 31% 24%

N55 12% 10% 18%

Years since graduation(mean [SD])

18.6 (81.4) 14.4 (9.9) 15.4 (11.3)

1-10 37% 39% 39%

11-20 36% 38% 33%

21-30 18% 16% 17%

N30 9% 7% 11%

Full time 85% 84% 85%

Solo 72% 66% 66%

1 site 89% 86% 91%

Member of national orstate association

64% 65% 63%

Patient care hoursper week

33.5 (11.5) 33.6 (9.8) 32.5 (10.3)

All practice hoursper week

35.8 (13.2) 36.1 (13.3) 36.6 (13.4)

Patient visits per week 106.4 (77.9) 110.2 (136.0) 106.4 (77.7)

722 Journal of Manipulative and Physiological TherapeuticsSmith and CarberNovember/December 2007Chiropractors as Safety Net Providers

Chiropractor DemographicsDemographics of mail survey respondents (Table 3)

reflect a largely male and white constituency, with mostengaged in full-time practice, in solo practice, and in a singlelocation. Many respondents (N50%) are younger than 45years, and most (N60%) graduated from chiropractic collegewithin the last 20 years.

Percentage of patient visits provided for reduced fee

Chiropractor Provision of Uncompensated “Safety Net” CareAs reported in Table 4, most DCs (N70%) report that they

do provide some form of uncompensated care to theirpatients in need. For DCs who do provide some form ofuncompensated care, approximately 6% of their practicetime and visits are provided at no charge, whereas more than12% of their practice visits and time may be dedicated toproviding care at a reduced charge to the patient.

DISCUSSION

The response rate achieved for our mailed survey ofchiropractors is comparable with that reported for othersimilar surveys of busy professionals.4-8 We used anappropriate methodological protocol that encouraged com-pliance and improved our response rate. Our phone follow-upprovided further information on nonrespondents, whichallowed us to calculate a more accurate effective responserate and to assess systematic bias in nonrespondents. It ispossible that nonrespondents to our survey of chiropractorsdisproportionately represented those chiropractors closer toretiring from practice, given that nonrespondents weresignificantly and substantially older, in practice much longer,and reported less hours spent in practice. Nonrespondents'higher propensity to being in solo practice and to not belongto a state or national professional associations may alsoindicate their propensity to disengagement from collectiveactions involving their profession. A propensity toward

nearing retirement and/or being disengaged could alsoreasonably predispose nonrespondents to a lack of interestin a survey such as ours, with cover letter requesting theirinput to guiding the future of the chiropractic profession inthe US health care system.

The response rate to our survey of chiropractors was verysimilar to that reported for the Community Tracking Study

13.4% 17.6% 14.5%

Page 6: Chiropractors as Safety Net Providers: First Report of Findings and Methods from a US Survey of Chiropractors

Practical Applications

This report documents the extent to which chiro-practic providers serve a vital role in the US healthcare safety net.Future research and policy analyses should includeconsideration of chiropractic care.

723Smith and CarberJournal of Manipulative and Physiological TherapeuticsChiropractors as Safety Net ProvidersVolume 30, Number 9

conducted by the Center for Health System Change (HSC-CTS).9 The HSC-CTS Survey of Physicians drew theirsample from master files of the American Medical Associa-tion and American Osteopathic Association and reportedHSC-CTS physician survey response rates ranging from52% to 65%.

As reported from HSC-CTS physician survey data, theproportion of physicians who provide uncompensatedcharity care has declined consistently over the past decade.10

In 1996 to 1997, 76% of US physicians provided some carefor free or reduced fee, whereas by 2004 to 2005, only 68%ofphysicians provided such charity care. In our survey ofchiropractors, conducted in 2002 to 2003, approximately70% of DCs provide some care for free, and more than 80%of DCs provide some care at reduced fee, although we cannotprovide trends on provision of chiropractic charity care overtime because ours was a single cross-sectional survey. Wesurveyed chiropractors' amount of charity care provided on aweekly basis (∼2 weekly hours provided for free and4 weekly hours at reduced fee), which also closelyapproximates or exceeds the amount of charity care reportedfrom the HSC-CTS physician survey that averaged 11 h/mofor those physicians who did provide charity care.10

The decline in physician charity care reported by HSC-CTS coincided with an increase in the number of uninsuredpeople over the same timeframe, further alarming policy-makers. Physicians providing free or reduced cost careto uninsured patients in their own practice or as volunteers atfree clinics have served a well-recognized important role inthe US health care safety net, and HSC-CTS investigatorsspeculate that charity care may decline as physiciansincreasingly face financial and time pressures, differentpractice arrangements, as well as changes in the medicalmarketplace.10,11

It should be noted that our query on uncompensatedchiropractic care replicated that query as posed in the HSC-CTS medical physician survey and specifically excludedservices provided for which the physician expected but didnot receive payment. Broadening the definition of uncom-pensated care to include both charity care as well as bad debtwould doubtlessly increase estimates of the volume andextent of such care that is routinely provided by health careproviders to uninsured or underinsured patients. Largeincreases in the number of uninsured has prompted arenewed effort at identifying effective policy approaches forexpanding health insurance coverage through fundamentalreforms of the nation's health insurance system, strengthen-ing of employer-based health insurance, or expansions ofexisting public insurance programs.12 For instance, expan-sion of existing public insurance could include elimination ofthe Medicare 2-year waiting period for people who aredisabled because it is known that the current waiting periodrequirement results in a severe cost burden to individuals andto providers in the form of charity care provision and thatdeferred health care can worsen conditions creating

additional health problems and higher cost.13 Future researchand policy analyses should include consideration ofchiropractic care because our report documents the extentto which chiropractic providers also serve a vital, althoughoften overlooked, role in this nation's health care safety net.

REFERENCES

1. Smith M, Carber L. Chiropractic health care in HealthProfessional Shortage Areas (HPSAs) of the U.S. Am J PublHealth 2002;92:2001-9.

2. Barclay S, Todd C, Finlay I, Grande G, Wyatt P. Not anotherquestionnaire! Maximizing the response rate, predicting non-response and assessing non-response bias in postal question-naire studies of GPs. Fam Pract 2002;19:105-11.

3. Edwards P, Roberts I, Clarke M, DiGuiseppi C, Pratap S, WentzR, Kwan I. Increasing response rates to postal questionnaires:systematic review. BMJ 2002;324:1183-91.

4. Asch DA, Jedrziewski MK, Christakis NA. Response rates tomail surveys published in medical journals. J Clin Epidemiol1997;50:1129-36.

5. Sudman S. Mail surveys of reluctant professionals. Eval Rev1985;9:349-60.

6. Roberts H, Pearson JC, Dengler R. Impact of a postcard versusa questionnaire as a first reminder in a postal lifestyle survey.J Epidemiol Community Health 1993;47:334-5.

7. RussellML,VerhoefMJ, InjeyanHS,McMorlandDG.Responserates for surveys of chiropractors. JMPT 2003;27:43-8.

8. Hay DA. A mail survey of health care professionals: analysis ofthe response. J Can Chiropr Assoc 1996;40:162-8.

9. Cunningham PJ, Tu HT. A changing picture of uncompensatedcare. Health Aff 1997;16:167-75.

10. Reed MC, Cunningham PJ, Stoddard J. Physicians pullingback from charity care. Center for Studying Health SystemChange Community Tracking Study; Issue Brief No. 42,August 2001. Available from Download: http://www.hschange.com/CONTENT/356/.

11. Cunningham PJ, May JH. A growing hole in the safety net:physician charity care declines again. Center for StudyingHealth System Change Community Tracking Study; TrackingReport No. 13, March 2006. Available from Download: http://www.hschange.com/CONTENT/826/.

12. Collins SR, Davis K, Kriss JL. An Analysis of LeadingCongressional Health Care Bills, 2005-2007: Part I InsuranceCoverage, The Commonwealth Fund; March 19, 2007.Available at: http://www.commonwealthfund.org/publications/publications_shav.htm?doc_id=469753.

13. Hayes RM, Beebe D, Kreamer H. Too Sick to Work, Too Soonfor Medicare: The Human Cost of the Two-Year MedicareWaiting Period for Americans with Disabilities, The Common-wealth Fund; April 11, 2007. Available at: http://www.commonwealthfund.org/publications/publications_shav.htm?doc_id=473514.

Page 7: Chiropractors as Safety Net Providers: First Report of Findings and Methods from a US Survey of Chiropractors

APPENDIX A. EXAMPLE OF SURVEY ITEMS

724 Journal of Manipulative and Physiological TherapeuticsSmith and CarberNovember/December 2007Chiropractors as Safety Net Providers

Page 8: Chiropractors as Safety Net Providers: First Report of Findings and Methods from a US Survey of Chiropractors

APPENDIX A (CONTINUED)

725Smith and CarberJournal of Manipulative and Physiological TherapeuticsChiropractors as Safety Net ProvidersVolume 30, Number 9

Page 9: Chiropractors as Safety Net Providers: First Report of Findings and Methods from a US Survey of Chiropractors

APPENDIX B. DATA ALGORITHMS AND CALCULATION OF RESPONSE RATES

726 Journal of Manipulative and Physiological TherapeuticsSmith and CarberNovember/December 2007Chiropractors as Safety Net Providers

Page 10: Chiropractors as Safety Net Providers: First Report of Findings and Methods from a US Survey of Chiropractors

APPENDIX B (CONTINUED)

727Smith and CarberJournal of Manipulative and Physiological TherapeuticsChiropractors as Safety Net ProvidersVolume 30, Number 9

Page 11: Chiropractors as Safety Net Providers: First Report of Findings and Methods from a US Survey of Chiropractors

CALCULATION OF RESPONSE RATES

The denominators were calculated as follows: MailSurvey Sent − INVALID, Reason “Not Found” (USPS BadAddress, No Yellow Page listing, No Answer after MultiplePhone Attempts) − INVALID, Reason “Other” (Not Practi-cing, Retired, Deceased)

The numerators were calculated in 2 ways:

Mail: Mail Survey Returned CompletedMail and phone: Mail Survey Returned Completed +Phone Follow-Up Completed

728 Journal of Manipulative and Physiological TherapeuticsSmith and CarberNovember/December 2007Chiropractors as Safety Net Providers