For all nine of the investigated PROMs, the majority of ICF codings were within the component “b - Body functions”, followed by the component “d - Activities and participation”. Table 3 outlines the number of ICF codings per level for all included measures. The component “b - Body functions” covered 100% of the content of the EORTC QLQ-C30 fatigue scale, followed by the PFS-12 (80%), the EORTC QLQ-FA12 (79%), the PROMIS Fatigue Item Bank (71%), the BFI (71%), the PRO-CTCAE (67%), EORTC CAT Core Fatigue (65%), the FACIT-Fatigue (60%) and the MFI-20 (50%). The component “d - Activities and participation” was most frequently coded for the FACIT-Fatigue (35%), the PRO-CTCAE (33%), the EORTC CAT Core Fatigue (31%), the PROMIS Fatigue Item Bank (29%), the BFI (29%), the PFS-12 (20%), the EORTC QLQ-FA12 (14%), and the MFI-20 (14%). In “e - Environmental factors” the EORTC QLQ-FA12 had 7% of its codings and the FACIT-Fatigue 5%. Content that was not definable in the ICF represented 4% of the codings of the EORTC CAT Core Fatigue and 36% of the MFI-20.

Regarding third-level categories, content of the EORTC CAT Core was coded most frequently in “b1300 Energy level (42%), “b4552 Fatiguability” (17%), and “d2301 Managing daily routine” (8%). All three items of the EORTC QLQ-C30 fatigue scale were exclusively linked to the third level category ‘b1300 Energy Level’. For the EORTC QLQ-FA12 the most frequent third-level categories were in “b1300 Energy level (43%), “b1528 Emotional functions, other specified” (17%), “d2301 Managing daily routine” and “b1608 Thought functions, other specified” (both 14%).

For the PROMIS Fatigue Item Bank the category “b1300 Energy level” covered 56% of the content, with all other third-level categories representing at most 4% of the content.

The FACIT-Fatigue consisted of content mostly coded in “b1300 Energy level” (35%), “b1528 Emotional functions, other specified” (15%), and “d2301 Managing daily routine” (15%).

For the BFI “b1300 Energy level” covered 59% of the codings, while all other used third-level categories had 1 coding (6%).

The MFI-20 included content related to “b1400 Sustaining attention” (18%), and “b1300 Energy level”, “b1301 Motivation”, and “b4552 Fatiguability” (all 9%).

Content of the PRO-CTCAE could be coded in only two third-level categories: “b1300 Energy level” (67%) and “d2301 Managing daily routine” (33%).

The PFS-12 consisted of content in “b1300 Energy level” (60%) and single codings (i.e., 7%) in six further categories. An overview of ICF codings assigned to each instrument is provided in Fig. 1, further details can be found in Supplementary Table S1.

Proportion of different codings from the International Classification of Functioning, Disability, and Health (ICF) assigned to each instument

Further, for the EORTC QLQ-FA12, the MFI-20, the PRO-CTCAE, and the PFS-12 the ICF categories were linked to each scale of these instruments. Details regarding the ICF coverage of each scale can be found in Supplementary Material (Table S2 and Figure S1).

The agreement between the raters at the second level was 135 out of 160 classifications (Inter-rater agreement of 84.4%).

Across all PROMs, half of all items were categorized as assessing general fatigue (78/160 items, 48.8%), while another 28.1% (45/160) were assigned to physical fatigue. Cognitive (21/160, 13.1%) and emotional fatigue (15/160, 9.8%) were assessed less frequently. One item (1/160, 0.6%) was categorized as ‘other’, since it does not ask about fatigue itself but about feeling understood by others (EORTC QLQ-FA12, item 12, ‘Did you feel that your tiredness is (was) not understood by the people who are close to you?’).

The 34 items of the EORTC CAT Core Fatigue item bank were assigned to physical (21/34, 61.8%) and general (13/34, 38.2%) fatigue, which is reflected in the EORTC QLQ-C30 fatigue scale as well (66.6% physical fatigue; and 33.3% general fatigue). The EORTC QLQ-FA12 assesses emotional (3/12, 25.0%) and cognitive (2/12, 16.7%) fatigue besides general (4/12, 33.3%) and physical (2/12, 16.7%) fatigue. Beyond that, the one item on feeling understood which was categorized as ‘other’ (1/12, 8.3%) takes a social issue related to fatigue into account.

The PROMIS item bank v1.0 for fatigue mostly covers general fatigue (29/54, 53.7%) but also has a focus on cognitive fatigue (13/54, 24.1%), which takes the second largest proportion followed by physical (9/54, 16.7%) and emotional (3/54, 5.6%) fatigue.

The MFI-20 has seven items on general fatigue (7/20, 35.0%), followed by five items on physical (5/20, 25.0%) and four on cognitive and emotional fatigue, respectively (each 4/20, 20.0%). The PFS-12 mainly covers general fatigue (7/12, 58.3%) as well, but also covers all other three components, i.e., physical, cognitive (each 2/12, 16.7%), and emotional (1/12, 8.3%).

The FACIT Fatigue Scale (v 4) and the BFI have no items on cognitive fatigue, but assess general (FACIT: 8/13, 61.5%; BFI: 7/10, 70.0%), physical (FACIT: 3/13, 23.1%; BFI: 1/10, 10.0%) and emotional (FACIT: 2/13, 15.4%; BFI: 2/10, 20.0%) aspects of fatigue.

Figure 2 displays the proportion of items per questionnaire that cover general, physical, emotional, and cognitive fatigue.

Items covering general, physical, cognitive, and emotional fatigue per instrument

Further, for the EORTC QLQ-FA12, the MFI-20, the PRO-CTCAE, and the PFS-12 the types of fatigue were classified for each scale of these instruments. Details regarding the scale-specific fatigue classification can be found in the Supplementary Material (Table S3 and Figure S2).

During classification, the reviewers initially agreed on 112 of 160 items, resulting an overall agreement of 70.0%. Most conflicts (42/48, 87.5%) occurred because one reviewer classified the item as general while the other chose one of the specific fatigue components.