Transition probabilities
| Annual probability | Probability Distribution* | Source | ||
| Transition | Undiagnosed (NMD−, SLD−, CC−) | Diagnosed (NMD+, SLD+, CC+) | ||
| NMD to SLD | Dependent on cycle number (online supplementary appendix, table 2.4a† and figure 2.1†), incorporating fibrosis progression rate of 0.12 (online supplementary appendix, table 2.4a†) | Undiagnosed adjusted by RR=0.67 (95% CI 0.21 to 2.07) (online supplementary appendix, table 2.4b†) | Log-normal (for RR); triangular for undiagnosed‡ | ref 23 ref 28 29 |
| SLD to CC | Undiagnosed adjusted by RR=0.63 (95% CI 0.06 to 6.45) (online supplementary appendix table, 2.4b†) | Log-normal (for RR); triangular for undiagnosed‡ | ||
| SLD to HCC | 0.4% | None | ref 50 | |
| CC to DC§ | CCI→DCIII: 7.3% CCI→DCIV: 1.3% CCII→DCIII: 28.5% CCII→DCIV: 8.5% | CCI→DCIII: 6.4% CCI→DCIV: 0.8% CCII→DCIII: 17.1% CCII→DCIV: 5.1% | Beta for diagnosed; log-normal for expert opinion multiplier | ref 27 and expert panel¶ |
| CC to HCC | 3.3% | 3% | Triangular for diagnosed; log-normal for expert opinion multiplier | ref 50 and expert panel¶ |
| NMD/SLD to death | Probability dependent on age | None | Office of National Statistics (ONS)23 51 | |
| CC to death§ | CCI→death: 10.2% CCII→death: 9.0% | CCI→death: 7.5% CCII→death: 6.6% | Beta for diagnosed; log-normal for expert opinion multiplier | ref 27 and expert panel¶ |
| DC to HCC | 3% | None | ref 50 | |
| DC to transplant | Age <70: 5%; age ≥70: 0% | Triangular | ref 50 | |
| DC to death | DCIII→death: 25.1% DCIV→death: 20.4% | Beta | ref 27 | |
| HCC to transplant | Age <65: 4%; age ≥65: 0% | None | ref 52 | |
| HCC to death | 53.0%/25.5%/17.2%/16.7% in first/second/third/fourth year 13.3% after fourth year | None | ref 53 | |
| Transplant to death | 16.6%/3.1%/3.1% in first/second/third year 2.9% after third year | None | ref 54 | |
*Probability distribution used in probabilistic sensitivity analysis.
†Online supplementary appendix 2.
‡Based on lower and upper limits of 95% CI for fibrosis progression rate (0.06 and 0.18), the minimal and maximal annual probabilities of progression for a given cycle number were generated (patterns analogical to those presented in figure 2.1 in online supplementary appendix 2). Then, triangular distribution was used to reflect uncertainty of probability estimate for a given cycle number.
§Where I, II, III and IV refer to the Baveno stages of cirrhosis.
¶Table 2.3 in online supplementary appendix 2.
CC, compensated cirrhosis; DC, decompensated cirrhosis; HCC, hepatocellular carcinoma; NMD, no/mild disease; ONS, Office of National Statistics; RR, relative risk; SLD, significant liver disease.