|
 |
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
© 2008 American Public Health Association
DOI: 10.2105/AJPH.2007.112185
RESEARCH AND PRACTICE |
Rui Li and Ping Zhang are with the Division of Diabetes Translation, Centers for Disease Control and Prevention, Atlanta, Ga. K.M. Venkat Narayan is with the Department of Global Health, School of Public Health, Emory University, Atlanta.
Correspondence: Requests for reprints should be sent to Rui Li, 4770 Buford Hwy NE, MS K-10, Atlanta, GA 30341 (e-mail: rli2{at}cdc.gov).
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONReferences |
|---|
Objectives. The Balanced Budget Act of 1997 authorized Medicare to expand the coverage of glucose monitors and strips to non–insulin users with diabetes and self-management training to non–hospital-based programs. We examined the impact of this expansion on self-monitoring of blood glucose among Medicare beneficiaries who were not using insulin to treat their diabetes.
Methods. With data from the 1996–2000 Behavioral Risk Factor Surveillance System and a logistic regression model using a complex survey design, we compared the probability of self-monitoring of blood glucose among Medicare beneficiaries at the frequency recommended by the American Academy of Family Physicians’ clinical guidelines before and after the Medicare expansion. We also compared the change in the frequency of self-monitoring of blood glucose during these periods between Medicare beneficiaries and persons with private insurance by using a difference-in-difference model.
Results. Medicare expansion was positively associated with the probability of self-monitoring of blood glucose for both Medicare beneficiaries and persons with private insurance; the magnitude was between 7.1 and 16.6 percentage points.
Conclusions. The Medicare expansion effectively increased the frequency of the recommended self-monitoring of blood glucose in the Medicare population.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONReferences |
|---|
Diabetes is a common, growing, and costly disease in the United States.1,2 A disproportionate burden of diabetes occurs among persons aged 65 and older, because of higher disease prevalence among this age group than among younger age groups. In 2004, the prevalence rate of diabetes was 18.1% among persons aged 65 to 74 years and 15.7% among persons aged 75 and older.3 During 2002, the annual direct medical costs for persons aged 65 and older ($47.6 billion) corresponded to an estimated 52% of the total direct medical costs for people with diabetes ($91.8 billion).4 Medicare, the insurer for people aged 65 years and older and for people of all ages who have certain disabilities or end-stage renal disease, bears a large proportion of the direct medical costs of diabetes. During 2005, 32% of the total Medicare expenditure was attributable to treating illness among persons with diabetes.5
Optimal glucose control can prevent or delay diabetes-related complications and thus may potentially reduce or postpone medical costs associated with treating these complications. Self-management education and the self-monitoring of blood glucose level are effective tools for achieving good glucose control.6–13 The federal government passed the Balanced Budget Act of 1997, effective on July 1, 1998, which expanded the Medicare benefits on diabetes. Before the expansion, Medicare covered blood glucose monitors and strips only for insulin users and covered outpatient self-management training only for hospital-based programs. The Balanced Budget Act expanded the coverage of glucose monitors and strips to non–insulin users, allowing 100 strips per 3 months, and expanded the coverage of self-management training to non–hospital-based programs. All coverage was subject to a 20% copayment.14–16 We evaluated the impact of the Medicare expansion on self-monitoring of blood glucose among beneficiaries who were not on insulin treatment, the population on which the policy change had the most direct impact.
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONReferences |
|---|
Study Population and Data
We used nationally representative data from the Behavioral Risk Factor Surveillance System, an annual, state-based, cross-sectional, random, land-based telephone survey of 150 000 to 210000 community-dwelling US adults.17 The Behavioral Risk Factor Surveillance System is composed of a core questionnaire and state-added modules. All states are required to use the core questionnaire, which includes questions about the respondent’s access to care, demographic characteristics, and personal health status and health behaviors. The state-added modules can vary by state and by year. Information collected from the diabetes modules for each respondent included age of diabetes onset, whether they were on insulin treatment, whether they had eye or foot complications, self-management behavior, and use of diabetes preventive services.17
We used data from 1996 to 2000 in our analysis to cover the periods both before and after the 1998 Medicare expansion took effect. Twenty-eight states surveyed with the diabetes module for all 5 years. The median response rate, based on the Council of American Survey Research Organizations, which considers the complex survey design and sampling methods of the Behavioral Risk Factor Surveillance System, ranged from 48.9% to 63.2% during these 5 years.18 Our study population included persons in these 28 states with Medicare or private insurance as their primary health insurance coverage who had self-reported diabetes and did not use insulin (n = 13419). The exclusion of those who had missing data at any of the variables (n = 3810), yielded a total of 9609 individuals in our final analysis. Overall, the group with missing data had more Blacks, fewer Whites, and more women than the group that had no missing data. Among Medicare beneficiaries, persons with missing data had a lower education level than did those without missing data.
Statistical Analysis
Definitions of the variables.
The primary outcome for our study was self-monitoring of blood glucose, because this was the main purpose of the Medicare-expanded coverage of the Balanced Budget Act.16 We obtained our primary outcome by asking the following question: "About how often do you check your blood for glucose or sugar?" The recommendation on the frequency of self-monitoring of blood glucose is at least 3 times a day for persons who are on insulin treatment; however, no consensus exists on recommended frequency for persons who are not on insulin treatment.8 The American Academy of Family Physicians recommends self-monitoring of blood glucose at least once a day for people who are on oral glucose agents only. No specific recommendation exists for people who are on diet control only. In Healthy People 2010, the national goal for self-monitoring of blood glucose is to increase the percentage of people with diabetes who perform self-monitoring of blood glucose at least once daily to 60%.19 Ideally, we should have examined the impact of the Medicare expansion by treatment mode. Unfortunately, the Behavioral Risk Factor Surveillance System only asked the question of whether a person took pills or was on diet control for diabetes starting in 2001. We used a dichotomous variable to identify whether a person performed self-monitoring of blood glucose at least once daily for all non–insulin users, no matter whether the patient was on oral glucose agents or only on diet control for diabetes treatment.
The Medicare expansion may have had different effects on self-monitoring of blood glucose at different time periods. Figure 1
shows the timeline of the Medicare expansion. The Balanced Budget Act was signed on August 1, 1997, but did not take effect until 11 months later, on July 1, 1998. However, some people may have responded to the pending law during the transition period (i.e., between the time that the Balanced Budget Act was passed and when the law actually took effect) because of their increased knowledge of self-monitoring of blood glucose through publicity of the Balanced Budget Act. Conversely, a sizable population may have required additional time to respond to the changed laws and regulations. We used a dichotomous variable to indicate before and after the Balanced Budget Act passed, and we further defined 5 categorical variables to capture the effect of the expansion at different time periods: (1) 13 to 19 months before the Balanced Budget Act, (2) 0 to 12 months before the Balanced Budget Act (baseline), (3) transitional period (after the Balanced Budget Act passed and before the law on diabetes coverage was effective), (4) 0 to 12 months after the law took effect, and (5) 13 to 30 months after the law took effect (see areas below the calendar line in Figure 1).
|
We estimated the impact of the Medicare expansion by controlling for the respondents’ duration of diabetes and age: persons with a longer duration of disease are more likely to be on oral medication and to have diabetes-related complications, and persons aged younger than 65 and on Medicare should have disabilities including diabetes-related disabilities, such as end-stage renal disease, to be eligible for Medicare. Among both groups, the frequency of self-monitoring of blood glucose should have been higher than among people with diabetes who did not meet these criteria. Other variables that we controlled for included patient’s gender, race, marital status, education, and family income, because these factors often affect the demand of health services. All control variables were coded as either categorical or dichotomous except for duration of diabetes, which was coded as a continuous variable.
Analysis
We examined the impact of the Medicare expansion using 2 different approaches. First, we conducted a pre- and postpolicy change analysis for the intended target population only—that is, people who had Medicare as their primary insurance. We used a multivariate logistic regression model with complex survey design. SUDAAN software version 8.0 (Research Triangle Institute, Research Triangle Park, NC) was used, which took into account both the Behavioral Risk Factor Surveillance System design effect such as strata, primary sampling units, and survey weights as well as multiple years of data. We reported predicted margins for the impact of the Medicare expansion. Marginal prediction method was used to calculate the predicted margins, which predicted the average probability of a person performing self-monitoring of blood glucose at different time periods during the Medicare expansion, while retaining the same demographic and social economic characteristics and other covariates.
Second, we used a difference-in-difference model to examine the impact of the Medicare expansion, because the policy’s effect in the pre- and postpolicy analysis could have been confounded by other environmental changes.20 Our comparison group in the model comprised people who had diabetes and were not treated with insulin but had private health insurance, provided by either their employers or themselves. In this model, if the confounding effect did exist, the "true" effect of the Medicare expansion would be equal to the change of the average probability of self-monitoring of blood glucose among the target population between the pre- and postpolicy period minus the change of the probability of self-monitoring of blood glucose of the comparison group during the same period.
To estimate the model, we first pooled the 2 groups (i.e., the policy target group and the comparison group) and added a group indicator variable to the model. Then we added the interaction terms between the group indicator and the time variables that were used to reflect the effect of the Medicare expansion at different time periods. The coefficients of the interaction terms demonstrated the direction and significance of the "true" policy effect. The magnitude of the Medicare expansion effect for a given period was the difference of the change of predicted margins of self-monitoring of blood glucose between the 2 groups during that period. The accumulative effect was the sum of the effects from all 3 periods if they were statistically significant. However, the private insurance comparison group may have been contaminated by the information about the importance of self-monitoring of blood glucose from the public promotion of the Medicare expansion. In this case, the change in the comparison group may have been a result of the contamination from the "treatment" effect.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONReferences |
|---|
Table 1
reports the weighted crude means and 95% confidence intervals (CIs) of the relevant variables. For Medicare non–insulin users, the mean duration of diabetes was 11 years; 16% were aged younger than 65, 36% had more than a high school education, and 80% had family incomes less than $35 000. Persons with private insurance were younger and were more likely to be male, White, and married, with a higher level of education and family income and with a shorter duration of diabetes than the Medicare beneficiaries.
|
Before the Balanced Budget Act passed, 22.3% of the Medicare and nearly 30% of the privately insured non–insulin users practiced self-monitoring of blood glucose at least once a day. After the Medicare expansion, this percentage increased to 35.1% among the Medicare population and to 35.2% for those with private health insurance. These differences are shown in Figure 2
.
|
Figure 3
shows the predicted margins of self-monitoring of blood glucose among Medicare beneficiaries for various time periods before and after Medicare expansion. The probability of performing self-monitoring of blood glucose 0 to 12 months before the Balanced Budget Act passed was not significantly different than the probability of performing self-monitoring 13 to 19 months before the Act. The probability of performing self-monitoring of blood glucose once a day increased from 23.5% during the 0 to 12 months before the Balanced Budget Act to 28.7% (P= .13) during the transition period, 32.9% (P< .01) during the first year after the Medicare expansion took effect, and 40.1% (P< .001) during the following 18 months. On average, compared with 1 year before the Balanced Budget Act, the probability of self-monitoring of blood glucose increased by 16.6 percentage points (40.1% minus 23.5%) for Medicare non–insulin users during the 30 months after the policy was signed.
|
For the non–insulin users covered by private insurance, the predicted probability of self-monitoring of blood glucose at least once daily was constant before the Balanced Budget Act passed and during the transitional period (Figure 3
). However, the probability increased from 30.3% during 1 year before the Balanced Budget Act passed to 36.7% (P= .07) during the first year after the Medicare expansion took effect, and to 36.8% (P= .04) during the following 18 months. Overall, the probability of self-monitoring among the privately insured group increased by 6.5 percentage points (36.8% minus 30.3%) during the 30 months after the policy was signed.
The coefficients of the interaction terms between the group indicators and the 3 time periods (transitional, 0–12 months after Medicare expansion, and 13–30 months after Medicare expansion) in the difference-in-difference model were all positive but were significant (P 
.01) only for the last period (data not shown). The difference in the change of probabilities of self-monitoring of blood glucose between the 2 groups during this last period was 7.1 percentage points: the change of predicted margins of self-monitoring of blood glucose in Medicare beneficiaries during the last period (40.1% – 32.9% = 7.2%) minus the change of predicted margins of self-monitoring of blood glucose in privately insured during the same period (36.8% – 36.7% = 0.1%).
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONReferences |
|---|
Laws and regulations are often used as policy tools for improving access to and quality of health care in the United States. They are frequently intended to improve access to those services that offer substantial health benefits but have been underutilized because of market failure or for other reasons such as unaffordability by the poor.21–24 The passing of the Balanced Budget Act was an important milestone in Medicare history. It emphasized the importance of prevention in improving the quality of chronic care and potentially controlling high medical costs. Through the Balanced Budget Act, Medicare expanded its coverage of diabetes monitors and strips and of self-management education. Subsequently, Medicare also expanded or mandated a series of diabetes preventive care services for persons with diabetes.25,26 However, whether this increase in coverage achieved its intended purposes is unknown. This study was the first to our knowledge that analyzed the effect of the Medicare expansion on the self-monitoring of blood glucose.
Our research showed that the Medicare expansion was positively associated with the likelihood of self-monitoring of blood glucose at least once a day for non–insulin users, the primary target population of the Medicare expansion. The effect was significant after the expansion took effect and continued increasing during the subsequent years. The overall increase of the probability of self-monitoring of blood glucose was 16.6 percentage points after the policy was signed.
Through the study of 10 managed care plans and 60 provider groups across the United States that served 180000 patients with diabetes, Karter et al.27 found that changing from a full out-of-pocket expense to a some out-of-pocket expense or to a no out-of-pocket expense increased the probability of self-monitoring of blood glucose by 9 percentage points and 16 percentage points, respectively. The Medicare expansion decreased out-of-pocket costs on monitors and strips from 100% to 20%. If Karter et al.’s results also apply to the Medicare population, the expansion would increase the probability of self-monitoring of blood glucose among non–insulin users by about 9 percentage points, or less than 16 percentage points. The higher response to changes in patients’ copayment from our study implies that persons with Medicare are more price-sensitive than are persons in health maintenance organizations. The higher response may also be attributed to the spillover effect from the publicity of the Balanced Budget Act as well as the expansion in self-management training.
Parente et al.28 reported that improving knowledge of the benefit of the preventive service had a substantial positive effect on the use of the preventive service among the elderly. Schade and McCombs29 found that Medicare beneficiaries who were more exposed to a mass media campaign showed higher use of diabetes services than those who were less exposed. If this was, in fact, the case, the greater effect of the Medicare expansion on self-monitoring of blood glucose than that in Karter’s study was the outcome of both (1) the decreased out-of-pocket costs on diabetes monitors and strips and (2) the increased knowledge about self-monitoring of blood glucose and benefit coverage.
Another possibility for the positive association reported between the Medicare expansion and the probability of performing self-monitoring of blood glucose among Medicare beneficiaries could be the secular trend. Since the late 1980s, after the publication of the results from several large clinical trials and the development of the chronic illness model,30–32 the importance of glycemic control and patient self-management in achieving diabetes treatment goals has been highlighted, which has led to the change in clinical guidelines, with self-management education and self-monitoring of blood glucose as integral parts of diabetes care.33,34 Expanded coverage of diabetes education and testing supplies in the 1997 Balanced Budget Act was a result of this clinical momentum.35
In addition, the requirement of reporting hemoglobin A1c (HbA1c) control by Health Plan Employer Data and Information Set may also provide incentives to health providers to encourage self-management behavior among patients with diabetes. Hence, the increased self-monitoring of blood glucose among Medicare beneficiaries after the Medicare benefit expansion might be partially attributed to the national trend of promoting self-management behavior to achieve optimal HbA1c control. Our results showed that during the post–Balanced Budget Act period, the probability of self-monitoring of blood glucose among the non–insulin users with private insurance also increased. If this increase was caused by national environmental factors, and the Medicare population responded the same, the Medicare expansion would be associated with a 7.1 percentage-point increase (the "true" policy effect we got from the difference-in-difference model) instead of a 16.6 percentage-point increase (results from the pre- and postpolicy analysis) in the probability of self-monitoring of blood glucose.
However, it is also possible that the increased probability of self-monitoring of blood glucose among persons with private insurance was caused by the spillover effect from the Medicare expansion. Persons with diabetes covered by private insurance might have gained better knowledge of the importance of self-monitoring of blood glucose because of the publicity surrounding the Balanced Budget Act. In addition, because Medicare holds an enormous market share in health care, Medicare coverage policies can greatly influence private sector activities. Employers may add the coverage of test supplies and self-management education to their insurance benefits following the Medicare expansion; the policy may also spark the proliferation of diabetes testing supply companies with aggressive marketing tools to influence patient behavior. The timing of the increased probability of self-monitoring of blood glucose among the privately insured appears to support the spillover hypothesis.
Figure 3 shows that the probability of self-monitoring of blood glucose among persons with private insurance kept constant before the expansion took effect and only had a 1-time increase during the time period of 0 to 12 months after the Medicare expansion was implemented. However, we don’t have enough information to differentiate the secular trend and the spillover effect. On the other hand, private insurance policies may have added diabetes-monitoring supplies for some other reasons specifically related to the privately insured; therefore, what the difference-in-difference model captured was not a national trend but changes only among private insurers. In this case, the difference-in-difference model would also underestimate the Medicare policy effect. Thus, we concluded that the Medicare expansion was associated with an increase in the probability of self-monitoring of blood glucose once daily among non–insulin users by a range of 7.1 to 16.6 percentage points.
Limitations
This study had several limitations. First, our results were subject to the limitations of the Behavioral Risk Factor Surveillance System survey design, such as the use of self-reported data. In addition, the Behavioral Risk Factor Surveillance System is a land telephone–based survey. Thus, households without land-based telephones, who are mostly people of low socioeconomic status and people with only cell phones, were underrepresented. Overall, however, the Behavioral Risk Factor Surveillance System is a reliable survey, and many studies have demonstrated its validity.36 Second, 30% of the respondents had missing data. After we deleted the family income variable, which had the most missing data from the regression equation, only 15% of respondents had missing data and the results did not change. Eleven percent of respondents did not respond to the dependent variable. The missing pattern seemed not to be correlated with family income. If the missing pattern were not random, our results would not apply to the people who did not respond to the question about self-monitoring of blood glucose. Third, the Behavioral Risk Factor Surveillance System is a cross-sectional survey; therefore, we could not make a causality inference. If there were confounding factors, our results might be biased.
Fourth, our results may have underestimated the Medicare expansion effect because: (1) many non–insulin users who manage their diabetes through diet do not need to perform self-monitoring of blood glucose as frequently as do persons who use diabetes medication, and therefore, the Medicare expansion may have had little effect on their self-management behavior; (2) the Medicare expansion may also have had little impact on Medicare beneficiaries who had secondary insurance coverage, such as private insurance or Medicaid, because their second insurance could have already covered diabetes monitors, strips, and self-management education; and (3) although our time frame incorporated the entire period during which the legislation was drafted, passed, and became effective, we have postpolicy data for only a 2.5-year period after the Medicare expansion took effect; thus, we are unable to capture the effect after our study period.
Fifth, in the difference-in-difference model, the privately insured may not be a good comparison group for the Medicare beneficiaries, because all the privately insured are younger than 65 and they may have different unmeasured characteristics than Medicare beneficiaries who consist of persons 65 and older and persons with disabilities. Finally, self-monitoring of blood glucose is only a process measure that facilitates diabetes glucose management. Ultimately, we would hope that this change in health coverage policies could impact the patient HbA1c level, diabetes complications, and costs related to poor glycemic control.
In conclusion, our results demonstrated that the Medicare expansion in 1998 was positively associated with the likelihood that diabetes patients performed self-monitoring of blood glucose as recommended. This result implies that laws and regulations can be effective public health policy tools to improve certain underutilized preventive diabetes care services. Although we analyzed a policy introduced in the late 1990s and for beneficiaries with diabetes only, our conclusion has important policy implications for current and future efforts to increase other underutilized preventive care services for persons with diabetes or other illnesses. Future studies are needed to examine the effect of the Medicare laws on the intermediate outcome measures such as HbA1c level, long-term health outcomes, and total Medicare health care expenditure.
|
Acknowledgments |
|---|
The authors thank 2 anonymous reviewers and Stephanie Rutledge at the Division of Diabetes Translation for their help on editing and for providing critical comments for the article.
Human Participant Protection
No protocol approval was needed for this study.
|
Footnotes |
|---|
Peer Reviewed
Contributions
R. Li originated the research question, designed the study, did the analysis, and led the writing. P. Zhang played a supervisory role and helped substantially on the study design and the writing of the article. K. M. Venkat Narayan helped with the study design, framing the article, and writing.
Accepted for publication May 21, 2007.
|
References |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONReferences |
|---|
1. Centers for Disease Control and Prevention. National diabetes fact sheet. 2005. Available at: http://www.cdc.gov/diabetes/pubs/estimates05.htm#prev2. Accessed January 30, 2007.
2. Narayan KM, Boyle JP, Geiss LS, Saaddine JB, Thompson TJ. Impact of recent increase on future diabetes burden: US, 2005–2050. Diabetes Care. 2006; 29:2114–2116.
3. Centers for Disease Control and Prevention. National diabetes surveillance system. Available at: http://www.cdc.gov/diabetes/statistics/prev/national/tablebyage.htm. Accessed January 31, 2007.
4. Hogan P, Dall T, Nikolov, P, American Diabetes Association. Economic costs of diabetes in the US in 2002. Diabetes Care. 2003;26:917–932.
5. Centers for Medicare and Medicaid. CMS acts to improve quality care for chronically ill beneficiaries, 2005 [press release]. Available at: http://www.cms.hhs.gov/apps/media/press/release.asp?counter=1521. Accessed January 30, 2007.
6. Saudek CD, Derr RL, Kalyani RR. Assessing glycemia in diabetes using self-monitoring blood glucose and hemoglobin A1c. JAMA. 2006;295:1688–1697.
7. Brownlee M, Hirsch IB. Glycemic variability: a hemoglobin A1c-independent risk factor for diabetic complications. JAMA. 2006;295:1707–1708.
8. Renard E. Monitoring glycemic control: the importance of self-monitoring of blood glucose. Am J Med. 2005;118(suppl 9A):12S–19S.[CrossRef][Web of Science][Medline]
9. Blonde L, Karter AJ. Current evidence regarding the value of self-monitored blood glucose testing. Am J Med. 2005;118(suppl 9A):20S–26S.[Web of Science][Medline]
10. Evans JM, Newton RW, Ruta DA, MacDonald TM, Stevenson RJ, Morris AD. Frequency of blood glucose monitoring in relation to glycaemic control: observational study with diabetes database. BMJ. 1999;319:83–86.
11. Gerich JE. The importance of tight glycemic control. Am J Med. 2005;118(suppl 9A):7S–11S.[Web of Science][Medline]
12. Karter AJ, Ackerson LM, Darbinian JA, et al. Self-monitoring of blood glucose levels and glycemic control: the Northern California Kaiser Permanente Diabetes registry. Am J Med. 2001;111:1–9.[Web of Science][Medline]
13. American Diabetes Association. Standards of medical care for patients with diabetes mellitus. Diabetes Care. 2002;25(suppl 1):S33–S49.[CrossRef]
14. Department of Health and Human Services, Assistant Secretary for Legislation. Testimony on 1998 Medicare Budget by Bruce C. Vladeck, Ph.D. 1997. Available at: http://www.hhs.gov/asl/testify/t970213c.html. Accessed January 31, 2007.
15. Illinois Department of Human Services. Medicare coverage of diabetes self-management education and blood glucose monitoring supplies. Available at: http://www.dhs.state.il.us/chp/ofh/FN/Diabetes/medicare.asp. Accessed January 31, 2007.
16. Centers for Medicare and Medicaid Services. New Medicare preventive benefits [press release]. 1997. Available at: http://www.cms.gov/apps/media/press/release.asp?Counter=332. Accessed January 31, 2007.
17. Centers for Disease Control and Prevention. Behavioral Risk Factor Surveillance System Web page. Available at: http://www.cdc.gov/brfss. Accessed January 31, 2007.
18. Centers for Disease Control and Prevention. 2000 Behavioral Risk Factor Surveillance System summary data quality report. Available at: ftp://ftp.cdc.gov/pub/Data/Brfss/2000SummaryDataQualityReport.pdf. Accessed January 31, 2007.
19. Objectives for diabetes. In: Healthy People 2010. Washington, DC: US Department of Health and Human Services; 2000. Available at: http://www.healthypeople.gov/Document/HTML/Volume1/05Diabetes.htm#_Toc494509747. Accessed January 31, 2007.
20. Wooldridge JM. Introductory Econometrics: A Modern Approach. Mason, Ohio: South-Western; 2003.
21. Liu Z, Dow WH, Norton EC. Effect of drive-through delivery laws on postpartum length of stay and hospital charges. J Health Econ. 2004;23:129–155.[CrossRef][Web of Science][Medline]
22. Sloan FA, Rattliff JR, Hall MA. Impacts of managed care patient protection laws on health services utilization and patient satisfaction with care. Health Serv Res. 2005;40:647–667.[CrossRef][Web of Science][Medline]
23. Edwards KM. State mandates and childhood immunization. JAMA. 2000;284:3171–3173.
24. Gruber J. The incidence of mandated maternity benefits. Am Econ Rev. 1994;84:622–641.[Web of Science][Medline]
25. Centers for Medicare and Medicaid Services. Diabetes self management overview. Available at: http://www.cms.hhs.gov/DiabetesSelfManagement. Accessed January 31, 2007.
26. Centers for Medicare and Medicaid Services. Diabetes screening overview. Available at: http://www.cms.hhs.gov/DiabetesScreening. Accessed January 31, 2007.
27. Karter AJ, Stevens MR, Herman WH, et al.; for the TRIAD Study Group. Out-of-pocket costs and diabetes preventive services: the Translating Research Into Action for Diabetes study. Diabetes Care. 2003;26:2294–2299.
28. Parente ST, Salkever DS, DaVanzo J. The role of consumer knowledge of insurance benefits in the demand for preventive health care among the elderly. Health Econ. 2005;14:25–38.[CrossRef][Web of Science][Medline]
29. Schade CP, McCombs M. Do mass media affect Medicare beneficiaries’ use of diabetes services? Am J Prev Med. 2005;29:51–53.[CrossRef][Web of Science][Medline]
30. The Diabetes Control and Complications Trial Research Group. The effect of intensive treatment of diabetes on the development and progression of long-term complications in insulin-dependent diabetes mellitus. New Engl J Med. 1989;329:977–986.[CrossRef]
31. UK Prospective Diabetes Study Group. Intensive blood-glucose control with sulphonylureas or insulin compared with conventional treatment and the risk of complications in patients with type 2 diabetes. Lancet. 1998;352:837–853.[CrossRef][Web of Science][Medline]
32. Wagner EH. Population-based management of diabetes care. Patient Educ Couns. 1995;226:225–230.
33. American Diabetes Association. Self-monitoring of blood glucose. Diabetes Care. 1994;17:81–86.[Web of Science][Medline]
34. The American Diabetes Association. National standards for diabetes self-management education programs. Task Force to Revise the National Standards. Diabetes Educ. 1995;21:189–190, 193.
35. Centers for Medicare and Medicaid Services. Medicare’s proposed regulation to implement new preventive services under Medicare Modernization Act [press release]. Available at: http://www.cms.hhs.gov/apps/media/press/release.asp?Counter=1135. Accessed April 20, 2007.
36. Centers for Disease Control and Prevention. Behavioral Risk Factor Surveillance System: methods, validity, and reliability bibliography. Available at: http://www.cdc.gov/brfss/pubs/mvr.htm. Accessed January 31, 2007.
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |