| | The Effect of a Hypertension Self-Management Intervention on Diabetes and Cholesterol ControlThis work was presented in part as a poster at the Society of General Internal Medicine Annual Meeting in Pittsburgh, Pennsylvania, April 9, 2008. Abstract BackgroundMost patient chronic disease self-management interventions target single-disease outcomes. We evaluated the effect of a tailored hypertension self-management intervention on the unintended targets of glycosylated hemoglobin (HbA1c) and low-density lipoprotein cholesterol (LDL-C). MethodsWe evaluated patients from the Veterans Study to Improve the Control of Hypertension, a 2-year randomized controlled trial. Patients received either a hypertension self-management intervention delivered by a nurse over the telephone or usual care. Although the study focused on hypertension self-management, we compared changes in HbA1c among a subgroup of 216 patients with diabetes and LDL-C among 528 patients with measurements during the study period. Changes in these laboratory values over time were compared between the 2 treatment groups using linear mixed-effects models. ConclusionsThere was a significant effect of the self-management intervention on the unintended target of HbA1c,but not LDL-C. Chronic disease self-management interventions might have “spill-over” effects on patients' comorbid chronic conditions. Cardiovascular disease remains the leading cause of death in the US, accounting for over one third of deaths.1 The modifiable risk factors of hypertension and hypercholesterolemia among the general population, and glycemic control among patients with diabetes, synergistically contribute to a patient's risk for cardiovascular events.2, 3 Control of these risk factors in combination substantially reduces a patient's risk for clinical events greater than any single risk factor alone.4 Despite the availability of effective therapy for diabetes, hypertension, and hypercholesterolemia, a study using the Third National Health and Nutrition Examination Survey found that only 7.3% of patients with diabetes simultaneously achieve recommended goals of therapy for all 3 conditions.5 Similarly, in the Veterans Affairs, only 4% of patients with diabetes achieve targets of simultaneous control of all 3 risk factors,6 and only 13% of patients with known cardiovascular disease achieve target blood pressure and cholesterol control.7 Clinical Significance•Although most self-management interventions are single-disease focused, there may be “spill-over” effects on related conditions that increase the overall effectiveness of self-management support. •Patients with diabetes who received a nurse telephone self-management intervention for blood pressure also had significant improvements in their HbA1c. •There was no effect of the self-management intervention on low-density lipoprotein cholesterol. The inability to achieve accepted targets of risk factor control likely arises from a complex interaction of patient and provider behaviors. Interventions that target self-management skills among patients with chronic diseases have been growing in popularity recently and have been suggested as an integral part of improving the quality of chronic disease care.8, 9 Self-management education goes beyond traditional patient education by seeking to motivate patient behavioral change, enhance patient confidence, and provide problem-solving skills to manage the day-to-day tasks in managing their chronic illness.10 Self-management education programs have sought to improve disease outcomes through improved adherence to medications, diet, and lifestyle; however, experts do not agree on what constitutes the essential elements of effective self-management education. A systematic review of chronic disease self-management programs in older adults reported modest but clinically important reductions of glycosolated hemoglobin (with an average reduction of 0.8%) and systolic blood pressure (average reduction of 5 mm Hg).11 Counseling and behavioral interventions to improve low-density lipoprotein cholesterol (LDL-C), however, have produced smaller benefits, with mean reductions ranging from 3 to 7 mg/dL.12, 13 Most self-management interventions target a single chronic disease and focus on disease-specific outcomes.11, 14, 15, 16, 17, 18, 19 The principles and methods of self-management support in chronic disease interventions are similar across disease states, suggesting that the benefits of a self-management intervention may extend beyond the intended targets.8 Therapeutic interventions that can improve multiple cardiovascular risk factors simultaneously would be particularly valuable in light of the growing prevalence of multimorbidity and synergistic relationship between risk factors. The Veterans' Study to Improve the Control of Hypertension (V-STITCH) was a randomized controlled trial that tested a patient-level self-management intervention delivered by a nurse over the phone and a provider-level computer decision support system to optimize medication management.20 The purpose of the present study is to evaluate the effect of the patient self-management intervention targeting patients' blood pressure on the unintended targets of diabetes and cholesterol control. Methods  Study Design The V-STITCH trial was a 2-year cluster randomized control trial. The interventions occurred at 2 levels: provider and patient. Primary care providers were first randomized to receive either the computer decision support system focusing on hypertension medication management delivered at the point of care during patient visits or usual care without the decision support interface. Within each participating provider's primary care panel, patients with hypertension were then randomized to receive either a nurse telephone self-management intervention or usual primary care. The patient was the unit of analysis, and the primary study outcome was blood pressure control defined by JNC 6 (“The Sixth Report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure”) criteria, which was the current guideline at the time of the study.21 The primary study results have been presented elsewhere and showed a modest effect of the self-management intervention on blood pressure control.22 There was no effect from the provider intervention on blood pressure control, nor was there a significant effect interaction between the patient and provider interventions. We therefore analyzed all patients receiving the telephone self-management intervention compared with all patients who did not receive this intervention. Participants The study was conducted in the Durham VA (Veterans Affairs) Medical Center primary care clinics (3 sites). We enrolled patients from 32 primary care providers (23 general internists, 7 physician assistants, and 2 registered nurse practitioners). To be eligible for the patient intervention, individuals were followed by one of the participating primary care providers; had a diagnosis of hypertension by an outpatient diagnostic code (International Classification of Diseases 401.0, 401.1, or 401.9); and had a filled prescription for hypertensive medication in the previous year. The research assistants sent letters and contacted 816 eligible patients by telephone before upcoming primary care appointments. Among these patients, 190 refused and 38 were excluded for not meeting inclusion criteria. We enrolled 588 individuals (76% recruitment rate) between March 2002 and April 2003. Patient randomization to the nurse self-management intervention occurred within each provider so that equal numbers of patients from each provider received the nurse self-management intervention or usual care. Study statisticians generated random assignment of interventions in blocks of 16, which were placed in sealed envelopes. Patients enrolled in the study were followed for 24 months from enrollment or until dropout. The study was approved by the Durham VA Medical Center (VAMC) Institutional Review Board and all patients provided written informed consent. For the present study, starting with the randomized sample of patients in V-STITCH, we created additional criteria to define appropriate cohorts of patients to evaluate the intervention effects on the outcomes of glycosolated hemoglobin (HbA1c) and LDL-C. First, to identify patients for inclusion in the analysis of HbA1c, we identified all study participants with a diagnosis of diabetes based either on self-report or an ICD-9 (International Classification of Diseases, 9th Revision) diagnosis code of 250.x. For the patients with diabetes enrolled in V-STITCH, we included all patients with laboratory measurements of HbA1c obtained from the Durham VAMC laboratory between 24 weeks before enrollment and 24 weeks after study completion. We obtained HbA1c values for 216 patients (99% of all patients with diabetes in the parent study). Similarly, to identify a cohort of patients to include in analysis of LDL-C, we queried the Durham VAMC medical records to identify patients enrolled in the V-STITCH study with LDL-C measurements obtained within 24 weeks before enrollment and 24 weeks after study completion. Of the 588 patients enrolled in V-STITCH, 528 patients had LDL-C measurements during this window through the VAMC laboratory (90%). Measures The two outcomes of interest for the present study were glycemic control measured by HbA1c, and fasting LDL cholesterol. All labs were obtained through the Durham VAMC clinical laboratory and abstracted from the electronic medical record. While some patients received care outside of the VA, we did not record any outside laboratory measurements. All LDL-C measurements were calculated from a fasting lipid panel using the Friedewald method.23 To describe the patient samples at the time of enrollment in the study, we defined the baseline HbA1c and LDL-C as the reading closest to enrollment that was within 24 weeks before or after this date. The following demographic information was collected by patient self-report at the time of enrollment: age, sex, race, marital status, education level, and financial status. Race was dichotomized as white or non-white. Patients who reported they had enough money to pay their bills only by cutting back on things or difficulty paying their bill no matter what was done were categorized as inadequate income.24 Participants were asked if they currently exercise or participate in an active physical sport. Similarly, patients were asked to report their current smoking status. The Rapid Estimate of Adult Literacy in Medicine was used to measure literacy.25 Intervention The nurse telephoned patients within 1 week of randomization and then every 2 months over 24 months to deliver the intervention. for a total of 12 nurse calls. At each call, the nurse delivered scripted information drawn from the following 9 educational and behavioral modules: hypertension knowledge; memory; social support; patient/provider communication; medication refill reminders; appointment compliance; health behaviors (diet, exercise, smoking, alcohol use); health literacy aids; and medication side effects. The information has been described in further detail in a separate report.26 To ensure that the intervention information was standardized, the nurse used a database application, which contained predetermined scripts and tailoring algorithms. There were no face-to-face meetings between the nurse and the patient. The majority of the patients in the intervention arm received all 12 phone calls (mean number of calls 11.0; median 12; range 3-12), and the average phone call lasted 5 minutes. Patients enrolled in both the intervention arm and usual care received routine primary care throughout the study from the Durham VAMC. Statistical Analyses Two separate analyses were conducted: one for patients with diabetes using HbA1c as the dependent variable, and one for all patients with LDL-C as the dependent variable. Because these outcomes were collected during the course of routine primary care, each patient had a varying number of HbA1c and LDL assessments at varying intervals of time. Thus, it is not possible to examine specific laboratory assessment values for all patients at set periods (eg, 6, 12, and 24 months), as is commonly done in longitudinal randomized controlled trials. We used a linear mixed-effects model to analyze changes in HbA1c and LDL over time27 because it is a method that can accommodate this type of data structure. Using this method, all patient HbA1c or LDL-C measurements obtained during the study time interval contributed to the estimation of a mean trajectory of laboratory values during the 2-year study period. Exploratory data analysis of HbA1c values indicated a linear trend over time in both the usual care and nurse telephone intervention groups, so time was defined as the number of weeks from baseline and treated as a linear, continuous variable in the mixed-effects model. The fixed effects in the HbA1c model included intervention group (nurse behavior vs. usual care), time in weeks, and intervention group by time interaction. For analysis of the LDL-C outcome, exploratory data analysis indicated a quadratic trend over time, with the average LDL-C initially increasing in both groups and then decreasing. To appropriately model the shape of this outcome distribution, we included the fixed effects of intervention group (nurse self-management vs. usual care), time in weeks, time squared, intervention group-by-time interaction, and intervention group-by-time2 interaction. In both the HbA1c and LDL-C models, to account for the correlation between a patient's repeated measurements over the study period, patient-level random effects in the form of a patient-level random intercept and slope also were included. The correlation between the random intercept and slope indicates how a patient's baseline level is related to their rate of 2-year change.28 There was no evidence of differential correlation structures for usual care as compared with intervention patients. For both outcomes, we report the estimated change in the laboratory assessment value between study enrollment (t =0 weeks) and study completion (t=104 weeks). Models for HbA1c and LDL were estimated using PROC MIXED in SAS Version 9.1 (SAS Institute Inc., Cary, NC). Results Results for Patients with Diabetes: HbA1c Of the 588 veterans enrolled in V-STITCH, there were 219 patients with diabetes (117 patients in the usual care arm and 102 patients in the intervention group). Of these, 216 patients had HbA1c measurements available for analysis and were included in the analysis. The Figure shows the patient flow from patient enrollment in the parent V-STITCH study, with the 2 cohorts identified for analysis in the present study. Patient characteristics for the patient sample with diabetes are shown in Table 1 and for the patient sample with LDL-C measurements in Table 2. There were no significant differences in any of the baseline patient characteristics between the 2 treatment arms. Changes in HbA1c over the course of intervention period are presented in Table 3. The mean HbA1c decreased by 0.28% in the intervention arm (95% confidence interval [CI], −0.59-+0.04), but increased by 0.18% in the usual care arm (95% CI −0.11-+0.47). Based on the linear mixed-effects model, the estimated mean reduction in HbA1c over 2 years in the intervention compared with the usual care group was 0.46% (95% CI, 0.04%-0.89%; P=.03). Patients' estimated baseline HbA1c (random intercept) and 2-year change (slope) were negatively correlated, indicating that patients with higher levels at baseline had steeper rates of improvement over the 2-year period. However, this correlation was similar in the intervention and control groups, suggesting that the intervention was not more effective in patients with higher baseline HbA1c. | ⁎ Estimated values based on linear mixed effects model. |
Results for Patients with Available Cholesterol Measurements: LDL-C Of the 588 patients enrolled in V-STITCH, 528 patients had LDL-C measurements within the defined time period and were included in the analysis (269 patients in the nurse self-management intervention and 259 in usual care). Patient characteristics are shown in Table 2 and are similar to the cohort examined for analysis of HbA1c outcome. There was no significant difference between the intervention and usual care for any baseline characteristics. Changes in LDL-C over the course of the intervention period are presented in Table 3. While LDL-C decreased over the 2-year study period in both groups, there was no significant between-group difference in this outcome, with an estimated mean difference of 0.9 mg/dL LDL-C lowering in the usual care group compared with the nurse self-management intervention (95% CI, −7.3-5.6; P=.79). Similar to the analysis of HbA1c, patients with higher LDL-C at baseline had steeper rates of improvement over the 2-year period; however, there was no differential effect between the intervention and control groups. Discussion In a randomized controlled trial of a nurse-administered tailored self-management intervention to improve blood pressure control, we evaluated the effect of the intervention on the unintended targets of diabetes and cholesterol control over 2 years. There was a modest difference in glycemic control between intervention and usual care patients that was statistically significant. There was no evidence of an intervention effect on LDL-C. There did not appear to be any significant difference in the intervention effect according to degree of baseline disease control. Although hypertension, diabetes, and dyslipidemia are pathophysiologically distinct, from the patient perspective they are similar in the self-management requirements for successful disease control. All 3 are usually asymptomatic, which has been reported as a barrier to effective self-management.29 The benefits of treatment are based on avoiding long-term future vascular events, and patients often lack accurate risk perception.30, 31 Patients can significantly improve disease control of all 3 through diet and lifestyle changes; however, such changes are often difficult to maintain.32 In addition, approximately 20%-50% of patients are nonadherent to chronic medications and suffer worse health outcomes as a result.33 Despite these similar self-management challenges, self-management interventions for hypertension, diabetes, or cholesterol typically focus on single disease-specific outcomes. Indeed, the present study is taken from a trial focused on changing patient behavior specifically within the context of hypertension, without focus on related cardiovascular risk factors. Because the patient behaviors required for effective hypertension care overlap with other cardiovascular risk factors, there is reason to believe that an intervention's effect may “spill over” to other related conditions. The findings of this study should be considered in light of prior self-management interventions directly targeting glycemic control or cholesterol. One useful way to compare the significance of intervention effects across studies and clinical outcomes is through standardized effect sizes. This measure divides the intervention effect by the standard deviation of the outcome variable. Other investigators have suggested an effect size of 0.2 as a “minimum clinically important difference,” with 0.5 representing a “moderate” effect and 0.8 a “large” effect.34 Using these criteria, our findings suggest a modest but clinically important effect size of −0.38 (−0.46 divided by the baseline standard deviation of 1.2) on the outcome HbA1c. This effect is comparable with a recent systematic review of self-management interventions for diabetes, which produced an average effect size of −0.36 on the outcome of HbA1c.11 Among self-management interventions targeting cholesterol in addition to other risk factors, the mean reduction in cholesterol is often much smaller, with effect sizes of <0.2.12, 13 It is interesting to note that our findings of indirect intervention effects on HbA1c and LDL-C are similar to that of studies targeting these outcomes. There are many possible explanations for our lack of observed treatment effect on cholesterol. Our dietary advice did not focus on saturated fat or cholesterol intake, which are the usual targets of dietary interventions. In addition, the patients were already starting with a low baseline LDL-C, and there were strong secular trends for more intensive cholesterol lowering after several large clinical trials and revised guidelines published in 2004.35 Even without these factors, other authors have found that the effect of self-management interventions varies according to disease type. One proposed explanation for this is that self-management is most effective when it includes ongoing disease monitoring by the patient that allows patients to respond to new information.18 While blood pressure and blood sugar are frequently monitored by patients and allow ongoing feedback on their progress, no such monitoring is available for lipid management, and patients must rely on the clinic-based assessments that are often separated by months or years. There are a few limitations of this study that should be considered when interpreting the results. First, our patient sample was almost entirely male and these findings may not generalize to female populations. In addition, HbA1c and LDL-C were not collected as part of the study protocol, and we relied on the measurement of these outcomes in the course of routine clinical care. To account for this, we used a model to estimate the average change in HbA1c and LDL-C over the study period rather than directly measure a difference between baseline and study completion. It also is important to note that we evaluated the intervention's effect on HbA1c and LDL-C over a 2-year period. Most self-management interventions for chronic disease last 6 months or less, and there appears to be a deterioration of intervention effects over longer time periods.18, 36 By choosing the longer time frame, we were able to evaluate lasting changes in disease control, although we were not able to determine if these changes persisted after completion of the intervention. The difficulty in maintaining behavioral change is one of the greatest challenges in chronic disease management, and further research is needed to determine if lasting changes in disease control can be sustained. In conclusion, there was some evidence that a telephone-administered nurse self-management intervention targeting blood pressure control may have a modest “spill-over” effect on diabetes control at 2 years; however, the intervention had no significant effect on LDL cholesterol. Given the growing prevalence of multimorbidity and the synergistic relationship between cardiovascular risk factors, interventions that can simultaneously target multiple determinants of risk may be particularly valuable in optimizing patient outcomes. Acknowledgements This research is supported by the Department of Veterans Affairs, Veterans Health Administration, Health Services Research and Development Service, investigator initiative grant 20-034. The first author was supported by Grant Number KL2 RR024127 from the National Center for Research Resources, a component of the National Institutes of Health (NIH), and NIH Roadmap for Medical Research. Its contents are solely the responsibility of the authors and do not necessarily represent the official view of the National Center for Research Resources, NIH, or the Department of Veterans Affairs. References 1. 1Rosamond W, Flegal K, Furie K, et al. Heart disease and stroke statistics—2008 update: a report from the American Heart Association Statistics Committee and Stroke Statistics Subcommittee. Circulation. 2008;117:e25–e146.
CrossRef
2. 2Jackson R, Lawes CM, Bennett DA, et al. Treatment with drugs to lower blood pressure and blood cholesterol based on an individual's absolute cardiovascular risk. Lancet. 2005;365(9457):434–441. Abstract | Full Text |
Full-Text PDF (103 KB)
|
CrossRef
3. 3Law MR, Wald NJ. Risk factor thresholds: their existence under scrutiny. BMJ. 2002;324(7353):1570–1576. 4. 4Goede P, Lund-Andersen H, Parving HH, Pedersen O. Effect of a multifactorial intervention on mortality in type 2 diabetes. N Engl J Med. 2008;358:580–591.
CrossRef
5. 5Saydah SH, Fradkin J, Cowie CC. Poor control of risk factors for vascular disease among adults with previously diagnosed diabetes. JAMA. 2004;291:335–342.
CrossRef
6. 6Jackson GL, Edelman D, Weinberger M. Simultaneous control of intermediate diabetes outcomes among Veterans Affairs primary care patients. J Gen Intern Med. 2006;21:1050–1056.
CrossRef
7. 7Johnson ML, Pietz K, Battleman DS, Beyth RJ. Therapeutic goal attainment in patients with hypertension and dyslipidemia. Med Care. 2006;44:39–46. MEDLINE |
CrossRef
8. 8Bodenheimer T, Lorig K, Holman H, Grumbach K. Patient self-management of chronic disease in primary care. JAMA. 2002;288:2469–2475. MEDLINE |
CrossRef
9. 9Institute of Medicine. Crossing the Quality Chasm: A New Health System for the 21st Century. Washington, DC: National Academy Press; 2001;. 10. 10Barlow J, Wright C, Sheasby J, et al. Self-management approaches for people with chronic conditions: a review. Patient Educ Couns. 2002;48:177–187. Abstract | Full Text |
Full-Text PDF (157 KB)
|
CrossRef
11. 11Chodosh J, Morton SC, Mojica W, et al. Meta-analysis: chronic disease self-management programs for older adults. Ann Intern Med. 2005;143:427–438. 12. 12Ebrahim S, Beswick A, Burke M, Davey Smith G. Multiple risk factor interventions for primary prevention of coronary heart disease. Cochrane Database Syst Rev. 2006;. 13. 13Brunner EJ, Rees K, Ward K, et al. Dietary advice for reducing cardiovascular risk. Cochrane Database Syst Rev. 2007;. 14. 14Ismail K, Winkley K, Rabe-Hesketh S. Systematic review and meta-analysis of randomised controlled trials of psychological interventions to improve glycaemic control in patients with type 2 diabetes. Lancet. 2004;363(9421):1589–1597. Abstract | Full Text |
Full-Text PDF (117 KB)
|
CrossRef
15. 15Schroeder K, Fahey T, Ebrahim S. How can we improve adherence to blood pressure-lowering medication in ambulatory care? (Systematic review of randomized controlled trials). Arch Intern Med. 2004;164:722–732. MEDLINE |
CrossRef
16. 16Norris SL, Lau J, Smith SJ, et al. Self-management education for adults with type 2 diabetes: a meta-analysis of the effect on glycemic control. Diabetes Care. 2002;25:1159–1171. MEDLINE |
CrossRef
17. 17Whelton SP, Chin A, Xin X, He J. Effect of aerobic exercise on blood pressure: a meta-analysis of randomized, controlled trials. Ann Intern Med. 2002;136:493–503. 18. 18Warsi A, Wang PS, LaValley MP, et al. Self-management education programs in chronic disease: a systematic review and methodological critique of the literature. Arch Intern Med. 2004;164:1641–1649. MEDLINE |
CrossRef
19. 19Newman S, Steed L, Mulligan K. Self-management interventions for chronic illness. Lancet. 2004;364(9444):1523–1537. Abstract | Full Text |
Full-Text PDF (225 KB)
|
CrossRef
20. 20Bosworth HB, Olsen MK, Goldstein MK, et al. The veterans' study to improve the control of hypertension (V-STITCH): design and methodology. Contemp Clin Trials. 2005;26:155–168. Abstract | Full Text |
Full-Text PDF (137 KB)
|
CrossRef
21. 21The sixth report of the Joint National Committee on prevention, detection, evaluation, and treatment of high blood pressure. Arch Intern Med. 1997;157:2413–2446. MEDLINE 22. 22Bosworth H, Olsen MK, Dudley T, et al. The Veterans' Study to Improve the Control of Hypertension (V-STITCH): A Patient and Provider Intervention to Improve Blood Pressure Control. Paper presented at 25th Health Services Research and Development national meeting, Washington, DC; 2007. 23. 23Friedewald WT, Levy RI, Fredrickson DS. Estimation of the concentration of low-density lipoprotein cholesterol in plasma, without use of the preparative ultracentrifuge. Clin Chem. 1972;18:499–502. MEDLINE 24. 24Bosworth HB, Bastian LA, Kuchibhatla MN, et al. Depressive symptoms, menopausal status, and climacteric symptoms in women at midlife. Psychosom Med. 2001;63:603–608. MEDLINE 25. 25Davis T, Long SW, Jackson RH, et al. Rapid estimate of adult literacy in medicine: a shortened screening instrument. Fam Med. 1993;25:391–395. MEDLINE 26. 26Bosworth HB, Olsen MK, Gentry P, et al. Nurse administered telephone intervention for blood pressure control: a patient-tailored multifactorial intervention. Patient Educ Couns. 2005;57:5–14. Abstract | Full Text |
Full-Text PDF (113 KB)
|
CrossRef
27. 27Diggle PJ, Heagerty P, Liang KY, Zeger SL. Analysis of Longitudinal Data. 2nd edn.. New York: Oxford University Press; 2002;. 28. 28Singer JD, Willett JB. Applied Longitudinal Data Analysis. New York: Oxford University Press; 2003;. 29. 29Jerant AF, von Friederichs-Fitzwater MM, Moore M. Patients' perceived barriers to active self-management of chronic conditions. Patient Educ Couns. 2005;57:300–307. Abstract | Full Text |
Full-Text PDF (133 KB)
|
CrossRef
30. 30Powers BJ, Oddone EZ, Grubber JM, et al. Perceived and actual stroke risk among men with hypertension. J Clin Hypertens (Greenwich). 2008;10:287–294.
CrossRef
31. 31Avis NE, Smith KW, McKinlay JB. Accuracy of perceptions of heart attack risk: what influences perceptions and can they be changed?. Am J Public Health. 1989;79:1608–1612. MEDLINE |
CrossRef
32. 32Dansinger ML, Tatsioni A, Wong JB, et al. Meta-analysis: the effect of dietary counseling for weight loss. Ann Intern Med. 2007;147(1):41–50. 33. 33DiMatteo MR, Giordani PJ, Lepper HS, Croghan TW. Patient adherence and medical treatment outcomes: a meta-analysis. Med Care. 2002;40:794–811. MEDLINE |
CrossRef
34. 34Samsa G, Edelman D, Rothman ML, et al. Determining clinically important differences in health status measures: a general approach with illustration to the Health Utilities Index Mark II. Pharmacoeconomics. 1999;15:141–155. MEDLINE |
CrossRef
35. 35Grundy SM, Cleeman JI, Merz CN, et al. Implications of recent clinical trials for the National Cholesterol Education Program Adult Treatment Panel III guidelines. Circulation. 2004;110:227–239.
CrossRef
36. 36Norris SL, Lau J, Smith SJ, et al. Self-management education for adults with type 2 diabetes: a meta-analysis of the effect on glycemic control. Diabetes Care. 2002;25:1159–1171. MEDLINE |
CrossRef
a Center for Health Services Research in Primary Care, Durham VA Medical Center, Durham, NC b Department of Medicine, Division of General Internal Medicine, Duke University, Durham, NC c Department of Biostatistics and Bioinformatics, Duke University, Durham, NC d Department of Psychiatry and Behavioral Sciences & Center for Aging and Human Development, Duke University, Durham, NC  Requests for reprints should be addressed to Benjamin J. Powers, MD, Center for Health Services Research in Primary Care, Hock Plaza, 2424 Erwin Road, Suite 1105, Durham, NC 27705
Funding: This research is supported by the Department of Veterans Affairs, Veterans Health Administration, Health Services Research and Development Service, investigator initiative grant 20-034. The first author was supported by Grant Number KL2 RR024127 from the National Center for Research Resources (NCRR), a component of the National Institutes of Health (NIH), and NIH Roadmap for Medical Research. Its contents are solely the responsibility of the authors and do not necessarily represent the official view of NCRR, NIH, or Department of Veterans Affairs. Conflict of Interest: None. Authorship: All authors had access to the data and contributed to writing the manuscript. PII: S0002-9343(09)00284-8 doi:10.1016/j.amjmed.2008.12.022 Published by Elsevier Inc. | |
|
FULL TEXT
