Table 3.
Synthetic result and related research scope of III linkage compared with DII linkage
| Analysis method | Study level of each topic* | Related research scope | ||||
|---|---|---|---|---|---|---|
| TC | GC | BC | PC | CC | ||
| Descriptive statistics | ○ | ○ | ○ | ○ | ○ | Epidemiologic indices including prevalence or cross-sectional proportion |
| Group comparison | ○ | ○ | ○ | ○ | ○ | Comparison of characteristics by the factor of interest |
| Incidence rate estimation | ○ | ○ | ○ | ○ | △ | Incidence and cumulative survival rates over time |
| Cox PH regression model | ○ | ○ | △ | ○ | △ | Evaluation of treatment effect, risk factor exploration, development of prediction model, etc. |
| Subgroup analysis | - | - | △ | X | - | Moderation effect evaluation by subgroup analysis or the interaction effect testing |
Abbreviations: III: indirectly identifiable information; DII: directly identifiable information; TC: thyroid cancer; GC: gastric cancer; BC: breast cancer; PC: prostate cancer; CC: cervical cancer; PH: proportional hazard
*Study level was divided into three categories: O = good, △ = insufficient, and X = poor