Health Affairs, 23, no. 4 (2004): 255-266 doi: 10.1377/hlthaff.23.4.255 © 2004 by Project HOPE	New Online   Pay Cuts For Medicare Docs   Access To Care Woes   Public Coverage More Efficient   Empowering Consumers

This Article Abstract Figures Only Reprint (PDF) Submit a response to this article Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Services E-mail this article to a friend Similar articles in this journal Similar articles in ISI Web of Science Similar articles in PubMed Alert me to new issues of the journal Add to My Personal Archive Download to Citation Manager Reprints & Permissions Citing Articles Citing Articles via HighWire Citing Articles via ISI Web of Science (20) Citing Articles via Google Scholar Google Scholar Articles by Villagra, V. G. Articles by Ahmed, T. Search for Related Content PubMed PubMed Citation Articles by Villagra, V. G. Articles by Ahmed, T. Related Collections Chronic Care Consumer Issues Health Promotion/Disease Prevention Health Reform Home Care Hospitals Insurance Coverage Managed Care Managed Care - Consumers Managed Care - Quality Quality Of Care

DataWatch

Effectiveness Of A Disease Management Program For Patients With Diabetes

Abstract

 
Diabetes disease management programs (DDMPs) are proliferating, but their effectiveness in improving quality and mitigating health care spending has been difficult to measure. Using two quasi-experimental methods, this study analyzed the first-year results of a multistate DDMP for people with diabetes sponsored by a national managed care organization. In both analyses, overall cost of care were significantly lower in DDMP sites, and the payer saved more than it spent. Pharmacy costs showed mixed results. Quality scores in the DDMP sites were significantly better than in sites without the program.

In spite of the widespread availability of evidence-supported information about effective treatments, many patients are not benefiting from them.⁴ Multiple studies have demonstrated that gaps in quality of care can be narrowed through comprehensive, multidisciplinary disease management programs (DMPs).⁵ Some programs can result in short-term decreases in use of acute care services and lower overall health care costs.⁶ Attracted by the positive impact on quality, use of services, medical costs, and worksite productivity, many employers, managed care organizations (MCOs), and insurance companies have adopted DMPs.⁷ Previous studies focused on economic outcomes and quality have included relatively few patients, have been confined to a single site or region, and therefore have limited generalizability.⁸ This paper evaluates the impact of a national MCO’s diabetes disease management program (DDMP) on quality of care, use of services, and costs in ten U.S. urban areas across twelve states.

Study Data And Methods

Top Study Data And Methods Study Results Discussion Editor's Notes NOTES

 
Study sites. We defined each study site as a unique administrative business unit of the sponsoring MCO under the jurisdiction of one or more health plans. The ten sites were Nashville, Tennessee; Florida; Denver, Colorado; the mid-Atlantic (Baltimore; Washington, D.C.; Philadelphia; Delaware; and southern New Jersey); Dallas, Texas; Houston, Texas; Chicago, Illinois; Kansas; Ohio; and Tri-State (New York metropolitan area, northern New Jersey, and Connecticut). The same DMP was delivered in all study sites. Site selection and implementation sequence were nonrandom, based on practical considerations such as customer request, stable staffing, and noninterference with previously scheduled National Committee for Quality Assurance (NCQA) accreditation activities. Program implementation was phased in over three years (Exhibit 1).

Program description. The DDMP was implemented as part of the general operation of a multisite MCO under usual business conditions rather than as a controlled experiment. Patients received assistance with self-care for diabetes and all common comorbidities such as cardiovascular and respiratory diseases. All educational contents were based on reputable national guidelines.⁹

All participants received graduated frequency and intensity of interventions according to a four-level severity stratification, with level four being those in need of most support and often the sickest. Stratification took place upon the first telephone contact with a nurse but was allowed to move up or down according to the patient’s health status and need for assistance at each subsequent assessment.

Interventions. For the purposes of this study, the first day of the one-year intervention period began two weeks after the first group of patients were mailed their "welcome package." The program consisted of repeated telephone outreach by trained nurses, dietitians, or health educators; Web-based education; remote monitoring devices; and reminders and educational mailings throughout the year. Physicians caring for participating members received information about the program prior to any contact with their patients and were notified of their progress by mail, fax, or telephone periodically thereafter. During the first nurse telephone call, patients underwent a structured interview that included a detailed lifestyle inventory and an assessment of the patient’s understanding of his or her disease(s) and adherence to prescribed drug treatments. Specific behavioral and cognitive goals were set for each patient and monitored regularly. Subsequent phone calls emphasized patients’ understanding of the American Diabetes Association (ADA) standards of care and other guidelines, adherence to their doctor’s treatment plans, motivation to adopt healthful lifestyle changes, and support to achieve those goals.

Because of the complex, multimodal nature of patient outreach activities, the intervention was deemed complete only after members had been exposed to the program for a minimum of ten months. This period, established a priori, was also considered necessary to give patients enough time to assimilate all of the information and modify their self-care behavior. In contrast, members with any length of program exposure (minimum of one month) are referred to as "all participants."

Data analyses. We conducted two separate analyses. The first analysis evaluated the program’s impact on quality, cost, and use among "full participants" (those enrolled in the first two months of the intervention period who remained in the program until the completion of the first year). All results were based on members’ entire one-year claim experience, with a three-month lag. The second analysis was based on "all participants" to approximate an intention-to-treat analysis.

Intervention and control groups. The three-year phased implementation schedule created "natural experiments" that allowed two sets of comparisons. The first was a sequential pre-post comparison in which each of the ten sites served as its own historical control. The "baseline" period was the year preceding the DDMP, and the "intervention" year corresponded to the first year of program operation. A second, parallel group comparison was also assembled by matching five DDMP (intervention) sites with a concurrent site in which no DDMP (control) was yet in place. When possible, sites were paired by geographic proximity.

Cost and use. The medical cost analysis counted each study member’s entire one-year claims experience regardless of the underlying diagnosis associated with each claim. In-network and out-of-network costs and use were counted. Encounters (services paid through capitation) were imputed with a local site fee-for-service equivalent value. Pharmacy costs were included, although some members had that benefit carved out. The proportion of such members remained stable throughout the study period. Claim caps and stop-loss rules were not applied. The costs of the DDMP itself are not reported because of the confidential nature of the contract between the MCO and the DDMP vendor, but we report the relative costs of the program against the results observed, as dollars per diabetic member per month.

Pre-post analysis. Several factors other than a DDMP influence year-over-year cost and use of services, and each site has unique market dynamics. These factors include provider contracts; changes in benefits; and unit price increases of drugs, devices, and other goods and services. These are usually not disease-specific and therefore tend to affect all medical costs. Adjustments for year-to-year overall medical inflation were made using each site’s cost trends. The reference population was the entire plan membership minus patients with diabetes. To account for possible shifts in demographics and comorbidity profiles related to member turnover, an age-sex and comorbidity adjustment was made using a multivariate member-month weighted regression method.¹⁰

Parallel group analysis. A similar case-mix adjustment was used when comparing concurrent intervention and control sites. In this case, the pre-post variable was switched to a control-intervention variable. Since three of the five site pairs were slightly off phase, an additional trend adjustment was made, so that intervention and control pairs were temporally aligned.¹¹

Quality of care. Quality was quantified using six indicators: dilated retinal exam, lipid testing, hemoglobin A1c (HbA1c) testing, microalbumin testing, prescribed angiotensin converting enzyme (ACE) inhibitors or angiotensin receptor blockers (ARBs), and tobacco use. The first four metrics were calculated according to the NCQA’s Health Plan Employer Data and Information Set (HEDIS) methodology.¹² Use of tobacco was based on patients’ self-reports during telephone visits. The metric represents the proportion of patients using tobacco regularly among all patients undergoing care by telephone. Assessment of the proportion of diabetic patients receiving a prescription for ACE inhibitors or ARBs was based on administrative data. Only the results among "full participants" are reported.

Study Results

Top Study Data And Methods Study Results Discussion Editor's Notes NOTES

 
Participation. A total of 55,439 members with diabetes were identified. Because a subset of companies opted not to cover the DDMP, 10,062 members (18.1 percent) were not eligible for the study. Another 1,885 eligible members (3.4 percent) opted out voluntarily or were falsely identified as diabetics, were not included in the program, and were not followed further. Of the remaining 43,492 members, 27,876 were included in the "full participant" pre-post analysis and 27,548 in the parallel group analysis. All 43,492 members were included in the "all participants" pre-post analysis, and 39,292 in the parallel group analysis. Exhibit 2 presents demographic characteristics, unadjusted baseline utilization patterns, and costs for both full participants and members excluded from the full participant analysis (n = 15,616, including patients who changed insurance carriers or died).

Physician feedback. All physicians were notified of their patients’ eligibility for the DDMP. Following initial patient contact by the DDMP staff, physicians received periodic updates of their patients’ progress. Although physicians’ input or feedback was sought, they rarely responded as intended in the program protocol. Because of this, few data were accrued regarding the impact of physician-DDMP-patient interaction on outcomes.

Cost and use, parallel group comparison. Results from all five site pairs were aggregated into intervention and control groups. Overall cost per diabetic member per month in the "full participants" intervention group was $417, compared with $554 in control sites (24.7 percent lower, p <.0001). Pharmacy costs in the intervention sites were $9.02 lower (–7.6 percent, p <.0002). Inpatient costs in the intervention sites were $17 (–11.4 percent) lower than in control sites (p <.08). There were 30 percent fewer admissions to the hospital (p <.0001) per 1,000 members in intervention sites than in control sites. Length-of-stay was 0.61 days higher in the intervention sites (11.6 percent, p <.0007). Results in the "all participants" analysis also showed an overall cost reduction (Exhibit 3).

View larger version (27K): [in this window] [in a new window]   EXHIBIT 5 Quality Indicators, Full Participants In The Parallel Group Analysis Of A Diabetes Disease Management Program (DDMP), 1997–2001

View larger version (27K): [in this window] [in a new window]   EXHIBIT 6 Quality Indicators, Full Participants In The Pre-Post Analysis Of A Diabetes Disease Management Program (DDMP), 1997–2001

Top Study Data And Methods Study Results Discussion Editor's Notes NOTES

Cost of care. The presence of the DDMP in any site was associated with significantly lower overall costs of care within one year. The most important source of savings was a 22–30 percent decrease in hospitalization. Savings measured using any length of program exposure were most likely lower because program effect was not measured at its mature performance point.

In the pre-post analysis, pharmacy costs were higher with the DDMP in place. This is not surprising, since the program actively promoted use of appropriate drugs and adherence to pharmacologic regimens. In the parallel group analysis, pharmacy costs were lower in the intervention group. The differences between these analytical approaches deserves further evaluation. The study was not designed to establish a direct cause-and-effect relationship between drug use and overall costs, but attempts have been made to establish such a link.¹³

Study limitations. The study’s reliance on quasi-experimental evaluation methods may elicit criticisms that the observations are subject to biases and confounders that could invalidate its conclusions. This is a fair criticism, but the cost and logistical challenge of a randomized controlled trial involving this many sites and patients would be formidable. However, the results of such trials often lack external validity, which could deprive patients of the benefits of effective therapies.¹⁴ Efforts were made to mitigate the most often cited objections to the quasi-experimental approach.¹⁵ All participants in the DDMP were taken into consideration regardless of level of disease severity. Excluding patients based on length of participation was a determination made before rather than a retrospectively applied criterion. Demographics and cost-and-use patterns did not differ significantly between the two groups. The pre-post analysis conferred some degree of intrasite stability. The analysis was strengthened by adjustment for age, sex, and comorbidity. This should mitigate concerns about year-to-year member turnover that could lead to shifts in case-mix between the baseline and intervention periods and imbalance between intervention and control sites in the parallel group analysis.

The use of redundant claims in case-finding queries to avoid false positives could lead to the selection of a sicker diabetic population than the total population with the disease. If that were the case, the reduction in cost in the pre-post analysis could be attributable to regression to the mean rather than program effect. Although this study design cannot fully eliminate regression to the mean, including patients of all severity levels in the analysis diminishes that concern. Furthermore, such a criticism would be valid only in the pre-post analysis but not in the parallel group analysis, where similar patient selection criteria applied to control and intervention sites. Another limitation of the study is the exclusion of voluntary opt-outs because data on these members were incomplete or missing.

Analysis of "all participants" approximates an intention-to-treat approach. This analysis does not reflect the actual conditions chosen for internal program evaluation because the program did not end at one year but continued beyond the end of the observation period of this study.

Impact on use. The impact of disease management on use of health services is of interest because disease management is driven entirely by patients’ and providers’ behavior changes, not by utilization controls. This is an attractive medical management model because of the backlash against utilization management tactics such as prior authorization for referral and hospital admissions. Recent evidence suggests that utilization control is, in fact, yielding to disease management as the preferred medical management strategy in a managed care setting.¹⁶ In this study, both analyses showed a decline in admission rates, increases in average length-of-stay, a net decline in bed days, and a slight decline in physician visits when a DDMP was in place. Lower admission rates and higher length-of-stay in intervention sites suggest more effective disease control in the nonhospital setting (including telephonic and other remote interactions between patients and caregivers). These results are similar to previous studies demonstrating decreases in hospital admissions associated with a DDMP.¹⁷

There was no a priori expectation of the DDMP’s impact on the number of out-patient visits. Both analyses showed fewer outpatient visits when the program was in place. These results are similar to those of previous studies.¹⁸ The explanation for decreased visits in our study is not readily apparent, but repeated telephonic and written communications between care managers and patients may have reduced patients’ need or desire to see a physician as often as they did before. In contrast to in-person visits, telephone contacts remove barriers stemming from lack of transportation, lack of time, payment of an office copayment, or having to take time off work. The use of care manager–initiated outbound calls also decreases the need to rely on patients to call in with problems. This is particularly important if patients choose not to "bother" their physician with seemingly unimportant concerns. Many potentially serious complications could be averted in the course of "routine" disease management interactions. On a population basis, this study shows no apparent adverse impact from this observed change in the number of office visits. If this observation is confirmed in other studies, the implications could be significant and may influence reimbursement policy for telephone, Internet, or other remote patient-provider interactions. Physicians could then increase same-day access for patients who really need it without any additional investment by physician practices. These implications would be important for the Medicare population in light of passage of the Medicare Prescription Drug, Improvement, and Modernization Act (MMA) of 2003, which includes a provision for disease management.¹⁹ Additional research is needed to confirm our observations and to understand the root causes and trade-offs between more remote supportive care and on-site care on the quality and use of various health services.

Patients’ acceptance of the DDMP. The ability through a DDMP to reach large numbers of patients quickly and cost-effectively has the potential to change the public health profile and cost of care for patients with diabetes at a level commensurate with its epidemic proportion. Rapid patient identification and engagement provided multiple opportunities for patient interactions. The incentive for rapid program deployment was driven in part by a contract between the MCO and the DDMP vendor that formalized program performance evaluation in one-year blocs. Typically, program installation was completed within three months.²⁰ Speedy implementation creates additional value in the quest for better quality and economic performance in heath services, a factor that is seldom considered in large-scale initiatives.

Impact on quality of care. HEDIS indicators are widely accepted as valid quality measures. In this study diabetes-related HEDIS and non-HEDIS metrics improved when a DDMP was in place. Improved HEDIS scores noted in the pre-post analysis could be interpreted as being secular changes related to "other" health plan activities aimed specifically at improving HEDIS scores instead of being attributed to the DDMP’s specific effects. Higher scores in DDMP sites in the parallel group analysis, however, could not be interpreted as such, because both intervention and control health plans had HEDIS improvement initiatives in place.

In summary, the DDMP resulted in improved quality of care, reduced use of hospital beds, and overall reduction in the total cost of care. If others obtain similar findings, the disease management model could have profound effects on the organization of care for people with chronic illnesses.

Editor's Notes

Top Study Data And Methods Study Results Discussion Editor's Notes NOTES

 
Victor Villagra (victor.villagra{at}snet.net) is president of Health and Technology Vector Inc. in Farmington, Connecticut. Tamim Ahmed is assistant vice president, research and analysis, at CIGNA HealthCare in Bloomfield, Connecticut.

This study was funded by CIGNA HealthCare.

NOTES

Top Study Data And Methods Study Results Discussion Editor's Notes NOTES

 

1. American Diabetes Association, "All About Diabetes," www.diabetes.org/about-diabetes.jsp (14 May 2004).
2. Lewin Group for the American Diabetes Association, "Economic Costs of Diabetes in the U.S. in 2002," Diabetes Care 26, no. 3 (2003): 917–932[Abstract/Free Full Text]
3. A.H. Mokdad et al., "Diabetes Trends in the U.S.: 1990–1998," Diabetes Care 23, no. 9 (2000): 1278–1283[Abstract/Free Full Text]; M.I. Harris et al., "Prevalence of Diabetes, Impaired Fasting Glucose, and Impaired Glucose Tolerance in U.S. Adults: The Third National Health and Nutrition Examination Survey, 1988–1994," Diabetes Care 21, no. 4 (1998): 518–524[Abstract]; and P.W. Wilson et al., "Overweight and Obesity as Determinants of Cardiovascular Risk: The Framingham Experience," Archives of Internal Medicine 162, no. 16 (2002): 1867–1872.[Abstract/Free Full Text]
4. J.B. Saaddine et al., "A Diabetes Report Card for the United States: Quality of Care in the 1990s," Annals of Internal Medicine 136, no. 8 (2002): 565–574.[Abstract/Free Full Text]
5. R.E. Aubert et al., "Nurse Case Management to Improve Glycemic Control in Diabetic Patients in a Health Maintenance Organization: A Randomized, Controlled Trial," Annals of Internal Medicine 129, no. 8 (1998): 605–612[Abstract/Free Full Text]; J. Sidorov et al., "Disease Management for Diabetes Mellitus: Impact on Hemoglobin A1c," American Journal of Managed Care 6, no. 11 (2000): 1217–1226[ISI][Medline]; and C.M. Renders et al., "Interventions to Improve the Management of Diabetes in Primary Care, Outpatient, and Community Settings: A Systematic Review," Diabetes Care 24, no. 10 (2001): 1821–1833.[Abstract/Free Full Text]
6. E.H. Wagner et al., "Effect of Improved Glycemic Control on Health Care Costs and Utilization," Journal of the American Medical Association 285, no. 2 (2001): 182–189[Abstract/Free Full Text]; and R.J. Rubin, K.A. Dietrich, and A.D. Hawk, "Clinical and Economic Impact of Implementing a Comprehensive Diabetes Management Program in Managed Care," Journal of Clinical Endocrinology and Metabolism 83, no. 8 (1998): 2635–2642.[Abstract/Free Full Text]
7. G. Ellrodt, "Evidence-Based Disease Management," Journal of the American Medical Association 278, no. 20 (1997): 1687–1692[Abstract]; and R. Christensen, "Disease Management Programs," Employee Benefit Research Institute Notes 23, no. 8 (2002): 1–4.
8. J. Sidorov et al., "Does Diabetes Disease Management Save Money and Improve Outcomes? A Report of Simultaneous Short-Term Savings and Quality Improvement Associated with a Health Maintenance Organization–Sponsored Disease Management Program among Patients Fulfilling Health Employer Data and Information Set Criteria," Diabetes Care 25, no. 4 (2002): 684–689.[Abstract/Free Full Text]
9. ADA, Clinical Practice Recommendations, 2002, care.diabetesjournals.org/content/vol25/suppl_1/index.shtml (18 May 2004); American College of Cardiology/American Heart Association, "Guidelines for Preventing Heart Attack and Death in Patients with Atherosclerotic Cardiovascular Disease: 2001 Update," Circulation 104, no. 13 (2001): 1527–1579; and American Thoracic Society, "Standards for the Diagnosis and Care of Patients with Chronic Obstructive Pulmonary Disease," American Journal of Respiratory and Critical Care Medicine 152, no. 5, Part 2 (1995): S77–S121.
10. Individuals’ average monthly cost and use outcomes measure were regressed by the following independent variables: (1) a binary variable representing pre-post time periods, (2) binary covariates for thirteen age/sex combinations (Exhibit 1), and (3) a comorbidity risk score based on disease episodes using a commercially available software (Episode Treatment Groups and Episode Risk Groups, Symmetry Health Data Systems Inc.).
11. For example, intervention site Houston and its control site Dallas overlapped by eleven months. To compensate and achieve twelve-month overlap, Houston’s cost was trended backward by one month.
12. National Committee for Quality Assurance, "NCQA HEDIS 2001 Data Submission," www.ncqa.org/programs/hedis/01sub.html (3 June 2003).
13. F.R. Lichtenberg, "Are the Benefits of Newer Drugs Worth Their Cost? Evidence from the 1996 MEPS," Health Affairs 20, no. 5 (2001): 241–251[Abstract/Free Full Text]; and S.B. Soumerai et al., "Effects of a Limit on Medicaid Drug-Reimbursement Benefits on the Use of Psychotropic Agents and Acute Mental Health Services by Patients with Schizophrenia," New England Journal of Medicine 331, no. 10 (1994): 650–655.[Abstract/Free Full Text]
14. R.K. Riegelman, Studying a Test and Testing a Test: How to Read the Medical Evidence, 4th ed. (Philadelphia: Lippincott Williams and Wilkins, 2000), 57, chap. 9.
15. A. Linden et al., "An Assessment of the Total Population Approach for Evaluating Disease Management Program Effectiveness," Disease Management 6, no. 2 (2003): 93–102.
16. S. Felt-Lisk and G.P. Mays, "Back to the Drawing Board: New Directions in Health Plans’ Care Management Strategies," Health Affairs 21, no. 5 (2002): 210–217.[Abstract/Free Full Text]
17. G.D. Berg, "Diabetes Disease Management in a Community-Based Setting," Managed Care 11, no. 6 (2002): 42, 45–50; D.K. McCulloch et al., "A Population-Based Approach to Diabetes Management in a Primary Care Setting: Early Results and Lessons Learned," Effective Clinical Practice 1, no. 1 (1998): 12–22; and H.I. Goldberg et al., "Evidence-Based Management: Using Serial Firm Trials to Improve Diabetes Care Quality," Joint Commission Journal of Quality Improvement 28, no. 4 (2002): 155–166.
18. M.B. Davidson et al., "Effect of a Pharmacist-Managed Diabetes Care Program in a Free Medical Clinic," American Journal of Medical Quality 15, no. 4 (2000): 137–142.[Abstract]
19. House Committee on Ways and Means, "Summary of Medicare Conference Agreement," 21 November 2003, waysandmeans.house.gov/media/pdf/healthdocs/confagreement.pdf (3 March 2004).
20. D. Ferris et al., "Maximizing Patient Engagement in Large Scale National DM Initiatives: A Case Study in Diabetes" (Presentation at the Disease Management Association of America conference, San Antonio, Texas, 24 November 2002).

                      What's this?

This article has been cited by other articles:

				D. C. Kendrick, D. Bu, E. Pan, and B. Middleton Crossing the Evidence Chasm: Building Evidence Bridges from Process Changes to Clinical Outcomes J. Am. Med. Inform. Assoc., May 1, 2007; 14(3): 329 - 339. [Abstract] [Full Text] [PDF]

				D. Bu, E. Pan, J. Walker, J. Adler-Milstein, D. Kendrick, J. M. Hook, C. M. Cusack, D. W. Bates, and B. Middleton Benefits of Information Technology-Enabled Diabetes Management Diabetes Care, May 1, 2007; 30(5): 1137 - 1142. [Abstract] [Full Text] [PDF]

				H. M. Krumholz, P. M. Currie, B. Riegel, C. O. Phillips, E. D. Peterson, R. Smith, C. W. Yancy, and D. P. Faxon A Taxonomy for Disease Management: A Scientific Statement From the American Heart Association Disease Management Taxonomy Writing Group Circulation, September 26, 2006; 114(13): 1432 - 1445. [Abstract] [Full Text] [PDF]

				J. F. Coughlin, J. E. Pope, and B. R. Leedle Jr. Old Age, New Technology, and Future Innovations in Disease Management and Home Health Care Home Health Care Management Practice, April 1, 2006; 18(3): 196 - 207. [Abstract] [PDF]

				J. Beich, D. P. Scanlon, J. Ulbrecht, E. W. Ford, and I. A. Ibrahim The Role of Disease Management in Pay-for-Performance Programs for Improving the Care of Chronically Ill Patients Med Care Res Rev, February 1, 2006; 63(1_suppl): 96S - 116S. [Abstract] [PDF]

				T. Bodenheimer and A. Fernandez High and Rising Health Care Costs. Part 4: Can Costs Be Controlled While Preserving Quality? Ann Intern Med, July 5, 2005; 143(1): 26 - 31. [Abstract] [Full Text] [PDF]

				F. J. Crosson and P. Madvig Does Population Management Of Chronic Disease Lead To Lower Costs Of Care? Health Aff., November 1, 2004; 23(6): 76 - 78. [Abstract] [Full Text] [PDF]