Screening for Type 2 Diabetes Mellitus: A Cost-Effectiveness Analysis

  1. Thomas J. Hoerger, PhD;
  2. Russell Harris, MD, MPH;
  3. Katherine A. Hicks, MS;
  4. Katrina Donahue, MD, MPH;
  5. Stephen Sorensen, PhD; and
  6. Michael Engelgau, MD
  1. From RTI International, Research Triangle Park, North Carolina; University of North Carolina at Chapel Hill, Chapel Hill, North Carolina; and Centers for Disease Control and Prevention, Atlanta, Georgia.
    1. Figure. *Diabetes onset; †Diabetes diagnosis by screening; ‡Diabetes diagnosed clinically.
      View larger version:
      Figure. *Diabetes onset; †Diabetes diagnosis by screening; ‡Diabetes diagnosed clinically. Progression of persons screened for diabetes.
    2. Appendix Figure 1. The model is used to follow the disease progression of all members of a cohort simultaneously on 5 different disease paths. At the end of any period, the cohort occupies one state on each of the disease paths. For the simulation, transitions between states take place at discrete time intervals 1 year apart. Thus, at the end of each 1-year period, portions of the cohort can move from one disease state to another or stay in the same disease state. The simulation program determines what proportion of the cohort will move from one state to another on the basis of the transition probability. In several cases, an individual can experience a complication event that the patient either dies of or survives during the period. The Markov model keeps track of the number of patients who are in each state in each period. It also tracks the cumulative incidence of patients who have undergone complication events, such as lower-extremity amputation ( ), angina, cardiac arrest ( ) or myocardial infarction ( ), and stroke. In the diagrams, complication events are represented by diamonds; states are numbered and represented by ovals. CHD = coronary heart disease. ESRD = end-stage renal disease.
      View larger version:
      Appendix Figure 1. The model is used to follow the disease progression of all members of a cohort simultaneously on 5 different disease paths. At the end of any period, the cohort occupies one state on each of the disease paths. For the simulation, transitions between states take place at discrete time intervals 1 year apart. Thus, at the end of each 1-year period, portions of the cohort can move from one disease state to another or stay in the same disease state. The simulation program determines what proportion of the cohort will move from one state to another on the basis of the transition probability. In several cases, an individual can experience a complication event that the patient either dies of or survives during the period. The Markov model keeps track of the number of patients who are in each state in each period. It also tracks the cumulative incidence of patients who have undergone complication events, such as lower-extremity amputation ( ), angina, cardiac arrest ( ) or myocardial infarction ( ), and stroke. In the diagrams, complication events are represented by diamonds; states are numbered and represented by ovals. CHD = coronary heart disease. ESRD = end-stage renal disease. Markov model of diabetes disease progression.LEACAMI
    3. Appendix Figure 2. A probabilistic sensitivity analysis was conducted in which 129 critical parameters were simultaneously varied over probability distributions on the basis of published 95% CIs or other reasonable ranges. A cost-effectiveness ratio was computed for each of 1000 iterations for both targeted and universal screening of people age 55 years. QALY = quality-adjusted life-year.
      View larger version:
      Appendix Figure 2. A probabilistic sensitivity analysis was conducted in which 129 critical parameters were simultaneously varied over probability distributions on the basis of published 95% CIs or other reasonable ranges. A cost-effectiveness ratio was computed for each of 1000 iterations for both targeted and universal screening of people age 55 years. QALY = quality-adjusted life-year. Histograms of cost-effectiveness ratios resulting from probabilistic sensitivity analyses based on targeted (top) and universal (bottom) screening.

    Summary for Patients

    « Previous | Next Article »Table of Contents

    This Article

    1. Ann Intern Med May 4, 2004 vol. 140 no. 9 689-699
    1. AbstractFREE
    2. » Figures
    3. Full Text
    4. Full Text (PDF)
      1. Summary for Patients

    Rapid Responses

    1. Submit a response
    2. No responses published

    Current Issue

    1. November 3, 2009, 151 (9)
    1. Alert me to current issues