Table 1—

Probabilities* in the model

ItemBaseline value(Range) or [95% CI]References
Natural course
 Prevalence12
  No retinopathy (NR)0.741[0.73–0.75]
  Nonproliferative retinopathy (NPDR)0.254[0.24–0.26]
  Proliferative retinopathy (PDR)0.005[0.0035–0.0072]
 Progression1214
  NR to NPDR0.065[0.05–0.08]
  NPDR to PDR0.116[0.11–0.13]
  NPDR to clinical significant macular edema (CSME)0.115[0.07–0.17]
  PDR to legal blindness with photocoagulation0.017[0.01–0.04]
  CSME to legal blindness with photocoagulation0.015[0.00–0.05]
  PDR to legal blindness without photocoagulation0.088[0.01–0.30]
  CSME to legal blindness without photocoagulation0.050[0.02–0.11]
 Mortality multipliers8, 17
  NR1.8[1.60–2.00]
  NPDR1.36[1.24–1.48]
  PDR1.76[1.64–1.88]
  CSME1.76[1.64–1.88]
Utility
 Legal blindness0.71[0.58–0.84]30
Characteristics of screening test
 Nonteleophthalmology (face-to-face examination)21, 22
  NR called NPDR0.05[0.04–0.06]
  NR called PDR0.003[0.0002–0.012]
  NPDR called NR0.22[0.21–0.23]
  NPDR called PDR0.02[0.01–0.04]
  PDR called NR0.02[0.01–0.04]
  PDR called NPDR0.03[0.02–0.04]
  Sensitivity of CSME0.82[0.61–0.94]
  Specificity of CSME0.79[0.61–0.91]
 Teleophthalmology (digital image evaluation)23, 24
  NR called NPDR0.04[0.02–0.06]
  NPDR0.10[0.06–0.15]
  NPDR called PDR0.004[0.00–0.01]
  PDR called NPDR0.19[0.04–0.48]
  Sensitivity of CSME0.88[0.69–0.97]
  Specificity of CSME0.94[0.89–0.97]
Assumptions
 Number of type 2 diabetic patients750(10–1000)
 Percent of examined patients with nonteleophthalmology0.25(0.0–1.0)
 Percent of examined patients with teleophthalmology0.75(0.0–1.0)
  • *

    * All probabilities represent annual transition probabilities in the Markov model.